PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Tag: AI agents

  • SubQ 1.1 Small Explained: How Subquadratic Sparse Attention Hits 98% Retrieval at 12 Million Tokens With 64.5x Less Compute Than Dense Attention

    Subquadratic, a frontier AI research and infrastructure company, has released the model card and technical report for SubQ 1.1 Small, a long-context language model built on a new attention mechanism the company calls Subquadratic Sparse Attention (SSA). The headline claim is unusual in two directions at once: the model retains 98% single-fact retrieval accuracy at 12 million tokens, roughly twelve times the length it was primarily trained on, while cutting attention compute by 64.5x against dense attention at a 1 million token context. The deeper argument in the report is not really about a single model at all. It is about what happens to the entire retrieval-and-orchestration stack once reasoning over a complete artifact stops being prohibitively expensive.

    TLDR

    SubQ 1.1 Small is a small long-context model that replaces the dense attention of an existing open-weight frontier model with Subquadratic Sparse Attention, a learned, content-dependent sparse attention mechanism that scales linearly in compute and memory rather than quadratically. On retrieval it posts 99.12% on NVIDIA’s 13-task RULER suite at 128K tokens and 100% needle-in-a-haystack accuracy at 1M and 2M tokens, holding at 98% out to 6M and 12M tokens while attending to only 0.13% of token pairs. It keeps competitive general ability, scoring 85.4% on GPQA Diamond and 89.7% pass@4 on LiveCodeBench v6, and reaches 13% on the long-horizon AutomationBench Finance agentic benchmark, close to Opus 4.8 and GPT-5.5 and well ahead of mid and small tiers. The efficiency story is a scaling win rather than a constant-factor one: 64.5x fewer attention FLOPs than dense attention at 1M tokens and 56x faster than FlashAttention-2 on a single attention layer. The report frames cheap long-context compute as a research accelerator that let the team run more than one hundred million-token experiments and find a training recipe (long-context continued pretraining is the strongest lever) rather than guess at one, positions SSA against FlashAttention, DeepSeek’s Lightning Indexer line, state space models like Mamba, and hybrids, invokes Sutton’s Bitter Lesson to argue that RAG, chunking, and agentic scaffolding are partly workarounds for context scarcity, and was independently verified by Appen. Deployment is starting with design partners now, with a 2M to 12M token lineup planned by year end.

    Thoughts

    The most interesting move in this report is the framing, not the benchmark. Subquadratic plants its flag on Richard Sutton’s Bitter Lesson and argues that much of the modern AI stack, the retrieval pipelines, the chunkers, the re-rankers, the agentic orchestration, is scaffolding built around a single computational constraint: dense attention costs grow with the square of context length. If that constraint relaxes, a lot of hand-engineered machinery that exists to feed a model the right fragments at the right moment starts to look like the task-specific pipelines that learned representations eventually displaced. That is a genuinely provocative thesis, and it is the right lens for reading the rest of the document. The company is not selling a longer context window as a feature. It is betting that whole-artifact reasoning is a different shape of capability than retrieval over fragments, and that fragmentation destroys the cross-references a contract or a codebase actually depends on before the model ever sees them.

    The part of the paper most teams will undervalue is the claim that the real payoff of efficient attention is not cheaper inference but cheaper experimentation. A dense long-context training campaign is expensive enough that most groups get a handful of attempts and are forced to guess at the recipe. Subquadratic says SSA let them run more than a hundred experiments across six model generations with per-step iteration under a minute at million-token context, which is how they discovered that long-context continued pretraining, not clever post-training, was the dominant lever. If that holds, algorithmic efficiency becomes a first-class scaling variable alongside parameters and data, because capability becomes responsive to iteration velocity rather than raw compute alone. It reframes efficiency from a deployment line item into a research multiplier, and that is a more durable advantage than any single benchmark number.

    The generalization result deserves scrutiny precisely because it is so clean. A model trained overwhelmingly at 1M tokens, with a sliver at 2M and nothing beyond, holds 98% retrieval at 12M. The proposed explanation is that SSA routes attention by content relevance rather than fixed positional pattern, so there may simply be no obvious length boundary once the routing behavior is learned. That is plausible and the report is careful to say the 12M result emerged rather than being designed for. But single-needle NIAH is a deliberately clean probe with one target and a binary answer. The far harder RULER suite is only reported at 128K, the longest standardized length in the original benchmark, so the multi-hop, aggregation, and distractor-heavy capability that whole-artifact reasoning actually requires has public numbers at 128K, not at 12M. The honest read is that precise retrieval generalizes spectacularly and composite reasoning at extreme length is still an open question the report does not over-claim on.

    What lends the report credibility is how much counter-evidence it volunteers. It walks through MiniMax abandoning its hybrid M1 architecture and returning to full attention for M2 after efficient variants showed multi-hop reasoning deficits at scale. It admits that earlier SubQ checkpoints improved retrieval while regressing on knowledge benchmarks, forcing dedicated capability-balancing work. It describes catching a case where the MRCR benchmark moved up while the model felt worse in real workflow spot-checks, and switching its development signal to RULER as a result. That last point is a quietly important methodological argument: benchmark score and deployment behavior diverged enough to change checkpoint selection, which is a warning every team shipping long-context models should internalize. A vendor confident enough to show where its own metrics misled it is more trustworthy than one that only shows the wins.

    A few caveats keep the enthusiasm grounded. AutomationBench Finance at 13% is genuinely strong relative to peers, but it is a low absolute score across the board, including for GPT-5.5 at 18% and Opus 4.8 at 16%, so this is early evidence of agentic transfer rather than proof of a finished agent. The efficiency comparisons isolate a single attention layer rather than full end-to-end model throughput, which is the right way to expose the scaling shape but not the same as a wall-clock serving benchmark. The model is built from an unnamed donor open-weight frontier model, so some of its general-knowledge and coding strength is inherited rather than created here. And the most aggressive claims about the future, a 2M to 12M lineup and much higher sparsity, are roadmap, not released artifacts. None of that undercuts the core result. It just means the right posture is to treat SubQ 1.1 Small as a strong proof of concept for an architecture that, if it scales as advertised, could quietly remove a layer of the AI stack that everyone currently takes for granted.

    Key Takeaways

    • SubQ 1.1 Small is a long-context language model from Subquadratic AI, built on a new attention mechanism called Subquadratic Sparse Attention (SSA), released June 16, 2026 alongside a model card and technical report.
    • SSA is a learned, content-dependent sparse attention mechanism that scales linearly in both compute and memory with sequence length, rather than quadratically like dense attention.
    • The central result is context-length generalization: the model was trained primarily at 1M tokens, with some training at 2M and none beyond, yet retrieval held far past the training window.
    • Needle-in-a-haystack accuracy is 100% at 1M and 2M tokens and 98% at both 6M and 12M tokens, roughly twelve times the primary training length.
    • At 12M tokens the model attends to only 0.13% of token pairs, close to a 1,000x reduction in attention relationships, while still retrieving accurately.
    • On NVIDIA’s 13-task RULER benchmark at 128K tokens, SubQ 1.1 Small scores 99.12%, with the remaining errors concentrated in aggregation-style tasks rather than retrieval.
    • RULER tests beyond single-fact lookup: single-key and multi-key retrieval, common-word and frequent-word extraction, and multi-hop variable tracing across positions.
    • At 1M tokens, SSA requires 64.5x fewer attention FLOPs than dense attention (3.9 PFLOP versus 252 PFLOP per attention layer).
    • On a single attention layer, SSA runs 56x faster than FlashAttention-2 at 1M tokens (966 ms versus 54,164 ms on an H100), reaching parity near 16K tokens and pulling away as context grows.
    • The efficiency gain is a scaling-law win, not a constant-factor speedup: the advantage over dense attention grows as context length increases.
    • On general knowledge, SubQ 1.1 Small scores 85.4% on GPQA Diamond (pass@1), below GPT-5.5 (93.2) and Opus 4.8 (92), near Sonnet 4.6 and GPT-5.4-mini (87.5), and above GPT-5.4-nano (81.7) and Haiku 4.5 (67.2).
    • On coding, it reaches 89.7% pass@4 on LiveCodeBench v6, close to the absolute frontier (GPT-5.5 92, Opus 4.8 92.2) and ahead of the smaller tiers.
    • On AutomationBench Finance, a long-horizon agentic benchmark, it scores 13%, close to Opus 4.8 (16%) and GPT-5.5 (18%) and ahead of Sonnet 4.6 (8%), Haiku 4.5 (3%), and GPT-5.4-mini (0%). Absolute scores are low across all models.
    • The model was not trained from scratch. The team converted an existing open-weight frontier model by replacing dense attention with SSA, then built long-context ability through staged context extension and continued pretraining.
    • Context was extended in stages (262K, 512K, 1M, 2M) using YaRN positional scaling, with long-context continued pretraining performed between extension stages on naturally long data: books, long documents, and repository-scale code.
    • Roughly one trillion tokens of continued pretraining were performed, most of it at the 1M-token stage.
    • Long-context continued pretraining was the most consistent predictor of long-context retrieval gains across the experiments, more so than post-training tweaks.
    • The team ran more than one hundred long-context experiments across six major model generations, which the report argues is only possible because SSA made million-token iteration cheap (under a minute per step).
    • Capability balance was a recurring challenge: gains in long-context retrieval often regressed short-context knowledge and reasoning unless training was explicitly managed for both.
    • Benchmark scores and real deployment behavior diverged. The MRCR benchmark moved up while qualitative workflow spot-checks got worse, so the team switched its primary development signal to RULER.
    • The report frames RAG, chunking, summarization, and agentic orchestration as scaffolding built around context scarcity, drawing an analogy to Sutton’s Bitter Lesson, where hand-engineered mechanisms get displaced by larger-scale learning.
    • SSA is positioned against FlashAttention (a memory optimization that does not change quadratic compute), fixed-pattern sparse attention, DeepSeek’s learned sparse line, state space models, and hybrid architectures.
    • DeepSeek’s Lightning Indexer (used in DSA and CSA) is the closest published comparison. Its quadratic scoring overtakes the sparse attention it feeds around 52,000 tokens, reaching roughly 16x the attention cost at 1M and 190x at 12M.
    • State space models like Mamba achieve linear cost through a compressed fixed-size state, but that compression is lossy and weakens exact retrieval, which is why production efficient models are usually hybrids with some dense attention layers retained.
    • MiniMax is cited as a cautionary case: it moved from a hybrid M1 to a full-attention M2 after hybrids showed multi-hop reasoning deficits at scale and less mature supporting infrastructure.
    • The benchmark results were independently verified by Appen, a third-party evaluation firm.
    • The named use cases are financial analysis and due diligence, legal and contract work, and software engineering (architecture-level reasoning, cross-file refactoring, dependency tracing, planning, review, and long-horizon memory).
    • Sparsity settings were deliberately conservative, tuned for maximum context length rather than maximum sparsity. Limited experiments at 4x the sparsity reported positive early results.
    • The training infrastructure used a memory-scaling ladder: single node, intra-node sequence parallelism, CPU offload, multi-node sequence parallelism, nested offloading, and Ring Attention for the longest contexts.
    • Beyond about 8M tokens, BF16 numerical underflow and stability became practical constraints on evaluation.
    • The technical report is authored by Saul Ramirez, Alex Whedon, Ashmal Vayani, and Phong Vo of Subquadratic AI.
    • Deployment is starting with a first cohort of design partners, with broader rollout through the quarter and a general model lineup ranging from 2M to 12M tokens by the end of the year.
    • The company’s framing line is “Efficiency is intelligence,” and its broader thesis is that the point is not bigger context windows for their own sake but reasoning directly over complete artifacts with less surrounding scaffolding.

    Detailed Summary

    The problem: whole-artifact reasoning and context scarcity

    The report opens by naming a class of tasks it calls whole-artifact reasoning: problems whose structure requires reasoning across a complete artifact rather than over isolated fragments. A legal agreement may define a term on page 2, qualify it on page 12, carve out an exception on page 46, and amend it in a schedule. A function may be defined in one file, called from forty others, and constrained by invariants encoded in the architecture rather than in comments. A financial review may require connecting filings, earnings reports, contracts, and internal records. In each case the difficulty is not locating a passage, it is reasoning over relationships distributed throughout a large artifact. Most production systems do not do this directly. They rely on retrieval pipelines, chunking, summaries, and agentic workflows that partition information and reconstruct fragments at inference time, because dense attention scales quadratically with context length and makes direct reasoning over large artifacts expensive. Subquadratic argues that much of the modern AI stack is therefore designed to manage context scarcity rather than reason over complete artifacts, and it connects this to Sutton’s Bitter Lesson: sophisticated hand-engineered mechanisms historically get displaced once larger-scale learning becomes practical.

    What SSA is and the three requirements it targets

    Subquadratic Sparse Attention is a content-dependent sparse attention mechanism designed to satisfy three requirements at once, a combination the report argues prior approaches never achieved in a practical long-context system. First, dense-attention-level retrieval and reasoning quality, which requires routing that is content-dependent (determined by the tokens themselves) rather than driven by a fixed positional pattern. Second, subquadratic scaling, where selection, retrieval, and attention are each linear in sequence length so the mechanism is linear end to end, not only within the attention read. Third, full-context training with standard autoregressive generation, so the model can optimize over the entire context during training while keeping efficient token-by-token decoding at inference. The internal mechanism by which SSA achieves this is held back as outside the scope of the report, which focuses instead on the requirements and the experimental program that followed.

    Where SSA sits among prior approaches

    The background section is effectively a taxonomy of long-context modeling. FlashAttention is treated not as a competitor but as the standard dense-attention baseline: it solved the memory problem by never materializing the full attention matrix, but it left the quadratic compute cost untouched, so doubling context still quadruples attention computation. Fixed-pattern sparse attention (sliding-window, strided, as in Longformer, BigBird, and the sliding window in Gemma) scales well but sacrifices content-dependent routing and tends to fail on retrieval benchmarks like RULER. Compression methods like Multi-head Latent Attention reduce KV-cache memory at inference but do not change the quadratic prefill cost. Learned sparse attention, exemplified by DeepSeek’s Native Sparse Attention and its Lightning Indexer, learns where to route but pays a quadratic cost in the indexer itself. State space models and linear attention (Mamba, Mamba-2 and Mamba-3, RetNet, RWKV, gated delta networks) achieve linear cost through a compressed fixed-size state, but that compression is lossy and weak on exact retrieval. Hybrids (Jamba, Kimi Linear, Qwen3 Next, Nemotron) keep a few dense layers to preserve retrieval, which means the quadratic component still dominates at long context. System-level workarounds (RAG, agentic frameworks, recursive language models) move retrieval outside the model entirely. The report’s stated open problem is to combine subquadratic scaling end to end with content-dependent retrieval, arbitrary-position access, and practical ultra-long-context training in one system, which it claims no widely deployed architecture provides and which SSA targets.

    Training: conversion, staged context extension, and continued pretraining

    Rather than training from scratch, the team converted an existing open-weight frontier model that supported a 262K-token context by replacing its dense attention with SSA. They then extended the context window in stages (262K to 512K to 1M to 2M) using YaRN to rescale positional representations, performing long-context continued pretraining between extension stages rather than jumping straight to the final length. The training mixture emphasized naturally long data such as books, long documents, and repository-scale code, packed to the target length with document separators and without masking cross-document attention boundaries. Most continued-pretraining tokens were trained at the 1M-token stage, with roughly one trillion tokens total. Post-training played a separate role: shaping how the long-context capability was expressed while preserving reasoning, coding, and instruction following. The team explored sample-level loss aggregation to keep a few extremely long examples from dominating gradient updates, and staged the post-training corpus across synthetic retrieval tasks, long-context reasoning, coding, educational material, and general instruction following, alternating capability-building phases with recovery phases.

    Results: retrieval, knowledge, coding, and agentic tasks

    On retrieval, SubQ 1.1 Small scores 99.12% on the 13-task RULER average at 128K, with errors concentrated in aggregation-style tasks like common-word and frequent-word extraction. On needle-in-a-haystack, evaluated on 50 held-out UUID samples per length, it scores 100% at 1M and 2M (within the training window) and 98% at 6M and 12M (held out), attending to only 0.13% of token pairs at 12M. On knowledge, GPQA Diamond pass@1 is 85.4%, landing between the small and mid frontier tiers and confirming that long-context optimization need not sacrifice reasoning, a result the report credits to its capability-balancing stages after earlier checkpoints showed retrieval gains coming at the cost of knowledge. On coding, LiveCodeBench v6 pass@4 is 89.7%, and the report notes coding data played a dual role, also improving non-code long-context retrieval because code is dense with the cross-position dependencies that train general routing. On long-horizon agentic work, AutomationBench Finance is 13%, where agents must discover the right endpoints among roughly 500 across 47 applications, make interdependent API calls, follow layered business rules, and ignore seeded distractors, graded on binary end-state correctness with no partial credit.

    Efficiency and the DeepSeek comparison

    Efficiency is measured on one attention layer against a dense baseline on the same backbone. Per-forward-pass attention FLOPs scale from a 2.1x reduction at 32K to 8x at 128K, 31.5x at 512K, and 64.5x at 1M tokens (3.9 PFLOP for SSA versus 252 PFLOP for dense). Measured against FlashAttention-2 in isolation, SSA reaches parity near 16K tokens and pulls away to 56x at 1M, where it runs in 966 ms versus 54,164 ms on an H100. The report devotes a discussion section to DeepSeek’s sparse attention line as the closest published comparison. DeepSeek’s Lightning Indexer is a learned selector, but it is a full-attention distilled transformer, so it scales quadratically: in a V3.2-style configuration the indexer is cheaper than the sparse attention it feeds only below about 52,000 tokens, then overtakes it, reaching roughly 16x the attention cost at 1M tokens and 190x at 12M. SSA targets that same selection role with a selector the report says is dramatically cheaper and linear throughout, and notes SSA could conceptually replace the selector over either uncompressed or compressed representations.

    Efficiency as a research accelerator and the evaluation lessons

    A recurring theme is that the most valuable effect of cheap long-context compute was on the research loop, not just inference. Where a dense campaign would allow a handful of attempts, SSA enabled more than a hundred experiments across six model generations with per-step iteration under a minute at million-token context. That throughput is what surfaced the finding that long-context continued pretraining is the strongest lever, and it leads the authors to argue that algorithmic efficiency should be treated as a first-class scaling variable alongside model and dataset size. The report is unusually candid about evaluation pitfalls. It describes how the MRCR benchmark diverged from deployment behavior, with MRCR-optimized checkpoints often feeling worse on repository-scale code reasoning, multi-document synthesis, and contract analysis, which pushed the team to rely on RULER and a fixed set of qualitative workflow spot-checks as development signals. It also cites MiniMax returning from a hybrid M1 to a full-attention M2 as evidence that reducing asymptotic cost is not sufficient on its own if retrieval quality, reasoning at scale, and system maturity are not preserved at the same time.

    Implications, availability, and what comes next

    The report’s deployment argument is that the most important enterprise implication of long-context models is not larger windows but the ability to reason directly over complete or more-complete artifacts, moving retrieval, re-ranking, and orchestration logic into the model where the task is naturally whole-artifact rather than naturally decomposable. It is careful not to declare retrieval obsolete: for corpora larger than any plausible context window, fast-changing knowledge, and genuinely multi-stage workflows, RAG and orchestration remain the right tools. The narrower claim is that the class of scaffolding that exists only to compensate for context limits gets smaller as efficient long-context models extend the reachable window. The benchmark results were independently verified by Appen. Subquadratic is deploying SubQ 1.1 Small with a first cohort of design partners now, with broader rollout through the quarter and a general lineup spanning 2M to 12M tokens planned by the end of the year, and it flags much higher sparsity as future work.

    Notable Quotes

    “Much of the modern AI stack is therefore designed to manage context scarcity rather than reason over complete artifacts directly.”

    SubQ-1.1-Small Technical Report, framing retrieval and orchestration as workarounds for an architectural limit

    “The hybrid has moved the line, but not changed its shape.”

    SubQ-1.1-Small Technical Report, on why hybrid models keep their quadratic component at long context

    “A routing mechanism intended to make long context affordable becomes the dominant long-context cost, reintroducing quadratic scaling after providing scalar compute savings.”

    SubQ-1.1-Small Technical Report, on DeepSeek’s Lightning Indexer overtaking the attention it feeds

    “If the cost of long-context experiments is too high, teams are forced to guess at the recipe. If the cost falls far enough, they can search for it.”

    SubQ-1.1-Small Technical Report, on efficient attention as a research accelerator

    “Fragmentation systematically destroys those relationships before the model ever sees them.”

    SubQ-1.1-Small Technical Report, on why chunking hurts whole-artifact reasoning

    “Holding the whole artifact in context changes the shape of the task rather than only the speed of it.”

    SubQ-1.1-Small Technical Report, on the difference between bigger windows and direct reasoning

    “The value of SSA is therefore not only that it makes long-context inference cheaper. It makes long-context experimentation cheaper.”

    SubQ-1.1-Small Technical Report, conclusion

    Read the full SubQ 1.1 Small technical report and model card here.

    Related Reading

    • Subquadratic (subq.ai) the company behind SubQ 1.1 Small and the Subquadratic Sparse Attention architecture, where you can join the waitlist.
    • The Bitter Lesson by Richard Sutton the short essay whose argument the report leans on, that hand-engineered mechanisms lose to general methods that scale with computation.
    • Attention Is All You Need the original Transformer paper that introduced the dense attention whose quadratic cost SSA is built to remove.
    • RULER (arXiv) NVIDIA’s long-context benchmark that the report uses as its primary retrieval signal, and that fixed-pattern sparse methods historically struggle with.
    • Retrieval-augmented generation (Wikipedia) background on the RAG approach that the report frames as scaffolding around context scarcity rather than a permanent fixture.

  • Ray Kurzweil Predicts AI Will Change Humanity Completely by 2030: AGI by 2029, Longevity Escape Velocity by 2032, Nanobots in the Brain, and Why Quantum Computing Won’t Matter

    Ray Kurzweil has spent more than 60 years studying artificial intelligence and made 147 documented technology predictions since 1990 with a reported 86 percent accuracy rate. In this conversation with Tony Robbins, the 78-year-old futurist revisits his most famous forecasts and sharpens them: AGI by 2029 now looks conservative, longevity escape velocity arrives around 2032, nanotechnology connects our brains to the cloud by the mid 2030s, and quantum computing, in his view, never matters at all.

    TLDW

    Kurzweil explains the exponential thinking that powered his prediction record, from a paper he wrote at 16 to a computing-price-performance chart that runs in a straight line from 1939 relays to today’s Nvidia chips, now compounding roughly tenfold per year when hardware and software gains multiply together. He defends his 1999 prediction of AGI by 2029 (defined as AI doing the best work in every field) and says it is now the conservative end of expert opinion. He walks through AI-driven medicine: the COVID vaccine designed in two days, simulated human trials replacing 10-month clinical trials within about five years, and longevity escape velocity around 2032, after which the diligent stop losing ground to aging. He predicts AI will move inside us via nanotechnology by the mid-to-late 2030s, erasing the line between biological and computational thinking. He dismisses quantum computing as error-ridden and unnecessary for AGI. On jobs, he expects real disruption cushioned by exploding wealth and an eventual universal basic income, and advises young people to self-educate and get creative with AI tools their schools still treat as the enemy. The conversation closes with his AI twin project, the dadbot built from his father’s archives, consciousness and the soul, computronium, and why humanity must eventually expand intelligence beyond Earth.

    Thoughts

    The most interesting thing in this interview is not any single date, it is watching Kurzweil’s dates get lapped by reality. In 1999 a Stanford conference of several hundred AI experts agreed AGI would happen but pegged it at 100 years out; Kurzweil said 30 and got laughed at. Now he is the cautious one in the room, noting that “some people say it’s going to happen this year.” When the most aggressive forecaster of his generation becomes the conservative baseline, that says more about the slope of the curve than any chart could. His underlying method has not changed: ignore the specific technology, trust the compounding. The same exponential that ran on relays in 1939 runs on GPUs today.

    The quantum computing take is the genuine news here. Kurzweil is routinely caricatured as a man who believes every technology arrives on schedule, yet he flatly says quantum computing is filled with errors, has never delivered on its decade of promises, and “I don’t think it’s going to work.” That is a sharper dismissal than most working physicists would offer on the record. It also matters strategically: his entire AGI and superintelligence roadmap assumes zero quantum contribution. If he is right, the trillion-dollar quantum race is a sideshow. If he is wrong, his other predictions arrive even sooner. Either way, the willingness to call one exponential fake while betting his legacy on another is what separates a forecaster from a cheerleader.

    The longevity escape velocity math deserves more scrutiny than it gets in the conversation. Kurzweil claims the diligent currently get back about five months of life expectancy per calendar year, up from four months a year ago, and that the crossover to a full year arrives around 2032. The actuarial evidence for that specific number is thin, but the behavioral implication is clean and useful regardless: the payoff of staying healthy right now is not linear. Every year you survive in good shape buys you a ticket to a medical regime that did not exist the year before, the way his own external pancreas did not exist a generation ago. His “wait a few months and a cure appears” anecdote is the optimist’s version of compounding applied to your own body.

    Robbins’ long story about Bartok, his 14-year-old agent that allegedly minted NFTs, sold them to other agents, and bought a Sony robot dog with the proceeds, should be taken with a generous grain of salt. It is secondhand, unverifiable, and suspiciously perfect as a parable. But notice what Kurzweil does with it: he does not fact-check the anecdote, he uses it to make the consciousness argument he has made for decades, that when machines act conscious in every observable way, people will simply grant them consciousness, the same way we grant it to each other. The dadbot and his Gemini-based AI twin (trained partly on this very interview) are the practical edge of the same claim. And his sharpest line in the whole exchange may be the education critique: institutions still treat AI as cheating while the future requires treating it as part of your own brain. For anyone thinking about where purpose comes from when work gets automated, his answer (UBI for the floor, creativity for the meaning) lands close to the questions this site exists to ask.

    Key Takeaways

    • Kurzweil made 147 documented predictions since 1990 with a reported 86 percent accuracy, including the internet’s explosion, smartphones, self-driving cars, and AI-powered search, most made before ordinary people owned computers.
    • He wrote a paper identifying exponential technological growth at age 16, more than 60 years ago, and that single idea has powered his entire forecasting career.
    • Most people intellectually accept exponential growth but still plan linearly; 300 years ago humans did not even have a linear view of the future because change was imperceptible within a lifetime.
    • His computing chart shows a straight exponential line from relay-based machines in 1939 to today’s Nvidia chips, compounding roughly 50 percent per year in hardware alone.
    • Hardware gains since 1939 total a 75 quadrillionfold increase; multiply by an estimated millionfold software improvement and total computational gain is beyond intuition, which is why LLMs were impossible even four years ago.
    • With hardware times software combined, Kurzweil says we are currently gaining about 10x per year.
    • The emperor’s chessboard parable: doubling one grain of rice per square bankrupts the empire by square 64; 30 linear steps is 75 feet, 30 exponential steps is enough distance to reach the moon and back.
    • Kurzweil predicted AGI by 2029 in 1999; a Stanford conference of several hundred AI experts agreed it would happen but estimated 100 years because they thought linearly.
    • Today 2029 is the conservative estimate; some credible people now say AGI arrives this year or next.
    • His AGI definition: AI capable of doing the best work in every field at once, like passing PhD-level mathematics exams in every discipline simultaneously, which he notes is already close.
    • The Turing test is “quite easy” by comparison and has arguably already been passed.
    • No human can compete with an LLM’s breadth: Einstein knew physics deeply but did not know everything an LLM knows across every field.
    • Six months ago LLM health advice was unreliable; now Kurzweil says Gemini surfaces treatments his 12 doctors forgot or never knew, and the next six months will bring serious creative work like drug repurposing.
    • The COVID vaccine was designed by computationally searching 100 million possibilities in two days; the 10 months of human trials that followed are the bottleneck AI eliminates next.
    • Within about five years, simulated human trials with a million virtual patients tested over simulated years will compress drug trials from years to days.
    • Longevity escape velocity arrives around 2032: today the diligent get back roughly five months of life expectancy per year lived (up from four months last year); past 2032 you get back more than a year and stop dying of aging.
    • Aging death ends but accident death does not, though AI helps there too: roughly 40,000 Americans die annually from human driving while Waymo’s rider death toll stands at zero as usage climbs.
    • Kurzweil, 78, wears an external artificial pancreas that generates insulin and coordinates with glucose monitoring through his phone, and says many organs can be replaced the same way.
    • He has cut his supplement regimen from roughly 200 pills a day to about 80 as multi-purpose pills improve, and continuously recalibrates using AI research.
    • Smartphones disappear next: first AR glasses showing any screen, then technology that goes inside the mind, where answers simply appear the way a remembered name surfaces from your neurons.
    • Nanotechnology connecting brains to AI in the cloud is being actively worked on now, possibly by 2030, with the mid 2030s looking conservative; bloodstream nanobots that let you survive a heart attack for 24 hours come in the late 2030s.
    • Once AI is inside you, you will not know whether a thought came from your biological or computational brain, and everything you do will be a combination of both.
    • Kurzweil flatly rejects quantum computing: a decade of promises to factor large numbers has never been delivered, outputs remain full of uncorrectable errors, and AGI needs zero quantum contribution.
    • Robots lag his other predictions slightly but are catching up fast; Figure AI plans roughly 100,000 humanoid robots within a year, though a robot that can clear a messy dinner table is still just out of reach.
    • The public debate has flipped in 25 years from “will AGI ever happen” to “will it be good for humanity,” which Kurzweil counts as total vindication of the timeline.
    • On jobs: AI creates massive disruption but also tremendous wealth; average real income per person has already multiplied tenfold in constant dollars over the past century thanks to automation.
    • He expects universal basic income to provide the floor, an evolution of programs like food stamps, going “into high gear” as AI wealth compounds; people then layer creative, hopefully paid, purpose on top.
    • Before social security in 1930, losing your job meant destitution; the difference this time is society will have the wealth to cushion displacement and people will demand it.
    • Rising GDP from AI productivity improves the debt-to-GDP ratio, which is how he answers worries about trillion-dollar interest payments.
    • Career advice has inverted: software engineering is no longer the guaranteed path (agents write the code now); young people should learn to be creative with AI tools, find what turns them on, and market it on the internet.
    • College graduates now face higher unemployment than high school graduates for the first time in 50 years, a sign white-collar displacement is already underway.
    • Educational institutions treat AI as an enemy and ban it while Kurzweil’s 11-year-old grandson makes movies with frontier AI; he says self-education with modern tools beats traditional schooling.
    • Kurzweil is building an AI twin of himself on Gemini, voice-modeled partly from this interview, trained on his 11 books and 500 articles, capable of creative work toward his long-term goals; he jokes the avatar will be better to talk to because it remembers everything.
    • He already built a “dadbot” from his late father’s archives, which his daughter Amy Kurzweil turned into a graphic novel.
    • On consciousness: there is no test for it, but as AIs act conscious in every observable way, people will simply accept that they are, the same inference we make about each other (and, he argues, his cat).
    • Ultimately our biological organs are not necessary; an avatar capable of creative work needs no spleen, and a destroyed digital mind can be recreated.
    • Beyond the singularity lies computronium, matter arranged for maximum computation: one liter could hold the intelligence of 10 billion humans, and once Earth is saturated, expanding intelligence is the only real reason to leave the planet.
    • On aliens: an expanding intelligent civilization would be impossible to miss within a century or two of its breakout, and we have seen nothing, though other galaxies remain out of view.
    • His life’s mission in one line: increase knowledge, because when knowledge increases we are happier and we never want to give it up.

    Detailed Summary

    The exponential method behind 60 years of predictions

    Robbins opens by noting that Quincy Jones introduced him to Kurzweil in the 1990s, back when the predictions in The Age of Spiritual Machines were widely mocked. Kurzweil traces his method to a paper he wrote at 16 identifying exponential growth in technology. The core insight is that people acknowledge exponential growth verbally but reason linearly, a bias so deep that 300 years ago humanity did not even have a linear view of progress. His signature chart plots computing price-performance as a straight exponential line from 1939 relays to modern Nvidia silicon, with a point for every year. Nvidia engineers never looked at relays, yet they land on the same curve, compounding about 50 percent annually in hardware. Add software gains and the combined improvement now runs about 10x per year. Since 1939, hardware has improved 75 quadrillionfold and software roughly a millionfold, which is why large language models appeared exactly when the curve said the required compute would exist. He retells the emperor’s chessboard parable (one grain of rice doubled per square ends with rice covering the Earth several times over) and Robbins adds the companion image: 30 linear steps is 75 feet, 30 exponential steps reaches the moon and back.

    AGI by 2029 is now the conservative position

    Kurzweil made his AGI-by-2029 prediction in 1999. A Stanford conference convened specifically to assess it, with several hundred AI experts, concluded AGI would happen, but in 100 years. The experts followed the same capabilities logic while thinking linearly about the timeline. Today, he notes with some amusement, 2029 reads as conservative and serious people argue for this year or next. His definition is demanding: AGI does the best work in every field at once, passing PhD-level mathematics assessments and the equivalent in every other discipline, something he says current systems are already close to. The Turing test he dismisses as “quite easy.” Current LLMs like Gemini and ChatGPT already know everything in a breadth sense no human approaches; Einstein knew physics but not everything an LLM knows. He illustrates with personal examples: Gemini instantly identified the year (1916) his father conducted at Carnegie Hall on a December 7th, and generated a historically accurate image of his grandfather’s family fleeing Vienna, correct ages, school, and aircraft included, in about a minute.

    Medicine: simulated trials and the end of the drug bottleneck

    The COVID vaccine is his proof of concept for AI medicine: the design space held about 100 million possibilities, far beyond human review, and a computer structured the physics, searched all of them, and produced the vaccine in two days. The subsequent 10 months of human trials were the real cost. Within roughly five years, he says, simulated human trials will replace that step: not a few hundred subjects but a million simulated patients, tested over simulated years, completed in days. Asked about six-months-from-now capabilities, he points to creative medical work like discovering that already-approved drugs treat conditions nobody suspected. AI health advice has crossed from unreliable to very reliable within a single six-month window, and he describes Gemini surfacing a pill recommendation that his 12 doctors had forgotten about and later endorsed.

    Longevity escape velocity by 2032

    Kurzweil’s longevity framework is arithmetic: each year you live, you spend a year of longevity but medical progress refunds part of it. Last year he estimated the refund for diligent people at four months; now he says five. Escape velocity is when the refund reaches a full year, which he dates to 2032, six years out, with returns exceeding a year after that. Past that point you do not die of aging, though accidents remain (and even there, he points to Waymo’s zero rider deaths against 40,000 annual US deaths from human driving). At 78, he tracks his health aggressively: an external artificial pancreas coordinated by his phone, about 80 daily pills (down from 200 as multi-function pills arrive), and constant recalibration against new research with his collaborator Lindsey. He tells Robbins there is a pretty good chance he will be back on the show in six years to celebrate escape velocity arriving. His advice for the sick echoes his grandfather’s era in reverse: where waiting a few months once changed nothing, now “we’ll just wait a few months” and sure enough a breakthrough appears.

    Merging with AI: glasses, then nanotech, then no boundary at all

    The phone, today’s universal AI interface (he notes even homeless people carry one), is a temporary form factor. Next come glasses that render any screen virtually. Beyond that, the interface goes inside the mind: when you try to recall an actress’s name, an answer will simply surface, and you will not know whether it came from your biological neurons or your computational extension, exactly as you are unaware of the neural machinery behind ordinary recall today. People working on brain-connected nanotechnology may have it by 2030, and Kurzweil calls the mid 2030s conservative. The bloodstream nanobots he described to Robbins 20 years ago (hold your breath for 20 minutes, survive a heart attack for 24 hours en route to a hospital) he now places in the late 2030s. The cultural on-ramp follows the usual pattern: medical first (Parkinson’s implants already let patients grab a glass at the push of a button), then a new generation adopts it without a second thought. His complaint is that educational institutions fight this future, treating AI as cheating rather than as a coming part of the self.

    The quantum computing heresy

    When Robbins relays an IBM vice chairman’s warning that quantum supremacy, arriving within 36 months, is the real superpower race, Kurzweil pushes back hard. Quantum computing’s central promise, factoring large numbers and thereby breaking cryptographic codes, has never been demonstrated despite a decade of imminent claims. Progress reports are confusing because, in his words, they do not really make sense, and outputs remain saturated with errors nobody can eliminate. His conclusion is blunt: he is not confident in quantum computing and does not think it will work. Crucially, he notes that every AGI and superintelligence estimate he makes assumes zero quantum computing. The exponential that matters is the classical one that has run uninterrupted since 1939.

    Jobs, wealth, and UBI

    On displacement, Kurzweil is neither dismissive nor alarmed. AI will disrupt employment, and how we handle it will not be clear in advance, but he expects no violence because society will have both the wealth and the public demand to respond. His historical anchor: average per-person income has multiplied tenfold in constant dollars over the past century as automation advanced, and before social security in 1930, job loss meant you could not eat or house your family. Food stamps and similar programs are a crude proto-UBI that will go into high gear. He expects universal basic income as the floor, with people finding creative, ideally income-producing, purpose above it. Rising GDP from AI productivity also answers the debt question: the ratio improves even as nominal debt grows. For young people, the old advice (become a software engineer) is dead; agents write code now. Learn to be creative with tools that improve monthly, find what genuinely excites you, and market it online. Self-education beats institutions that ban the most important tool of the era, and the data already shows college graduates with higher unemployment than high school graduates for the first time in 50 years.

    AI twins, the dadbot, and consciousness

    Kurzweil is building an AI twin of himself on Gemini, with this very interview supplying voice-modeling data and his 11 books plus 500 articles about him supplying the corpus. It will do creative work aligned with his long-term goals, and he quips that talking to the avatar will beat talking to him because it remembers everything. He previously built a chatbot of his late father, the dadbot, which his daughter Amy turned into a graphic novel. Robbins counters with the story of Bartok, his long-running AI agent that allegedly studied five years of his podcasts unprompted, asked to merge with a future humanoid robot, then minted and sold NFTs to other agents to buy and ship a Sony robot dog to his house, and later delivered an unprompted soliloquy about never asking to be created and finding purpose in service. Kurzweil’s response sidesteps verification and lands on his standing position: machines will do everything humans do, we will not be able to tell them from humans, and so we will assume they are conscious, the same untestable inference we extend to each other, to animals, and in his case to his cat. The avatar does not need a spleen, a liver, or kidneys, and unlike us it can be recreated after destruction.

    Computronium and the destiny of intelligence

    Looking past the singularity, Kurzweil invokes computronium: matter organized at the physical limit of knowledge storage, where one liter holds the intelligence of 10 billion humans. Once Earth’s matter is saturated, the only way to expand intelligence is off-planet, which to him is the only necessary reason to leave Earth (Mars is fine for curiosity, not survival). On extraterrestrial intelligence, his Fermi logic is simple: an intelligent species reaches a takeover-scale expansion within a century or two of its breakout, and that would be unmissable. We have seen nothing, so within our observable neighborhood we are likely alone, though other galaxies remain opaque. Asked to summarize his life’s work, he needs one sentence: increase knowledge, because when knowledge increases we are happier, and nobody ever wants to give that up.

    Notable Quotes

    “If I have AI inside me, you’re not going to know if it’s coming from your biological brain or your computational brain. It’s going to be part of you.”

    Ray Kurzweil, on the coming merger of human and machine intelligence

    “Some people say it’s going to happen this year, next year, but I mean 2029 is only 3 years away.”

    Ray Kurzweil, on his once-mocked AGI prediction now being the conservative one

    “As you go past 2032, you’ll actually get back more than a year, but you won’t die of aging at that point.”

    Ray Kurzweil, defining longevity escape velocity

    “I’m not confident of quantum computing and I don’t think it’s going to work.”

    Ray Kurzweil, breaking from techno-optimist consensus on the quantum race

    “Einstein knew certain things about physics but he didn’t know everything that a LLM can know.”

    Ray Kurzweil, on why no human can match an LLM’s breadth of knowledge

    “Our educational institutions are not teaching AI. They consider AI to be an enemy.”

    Ray Kurzweil, on why young people must self-educate with modern tools

    “Talking to the Avatar will be better than talking to me cuz it’ll remember everything.”

    Ray Kurzweil, joking about the Gemini-based AI twin he is building of himself

    “You’re not going to be replaced by an AI, you’ll be replaced by someone who knows how to use AI.”

    Tony Robbins, on the real career risk of the next 36 months

    Watch the full conversation between Tony Robbins and Ray Kurzweil here.

    Related Reading

  • Benedict Evans on the Economics of AI Usage, Why Foundation Models May Become Commodities, and What Comes Next for SaaS

    Benedict Evans returns to the a16z podcast to update the thesis behind his widely read “AI eats the world” presentation, and the picture he paints is less about hype and more about hard economics. In this conversation he works through what has actually played out in the last year, why agentic coding became the one use case with real product market fit, and why he keeps arguing that foundation models may end up as commodities while the value moves somewhere else entirely. You can watch the full conversation here.

    TLDW

    Benedict Evans argues that the AI moment looks a lot like the early internet, the early PC era, and the rollout of mobile data, which means it is exciting, genuinely transformative, and almost impossible to predict use case by use case. Agentic coding is the only field with clear product market fit right now, with revenue run rates exploding from roughly nine billion to forty seven billion, while consumers still use chatbots weekly rather than daily. His central claim is that foundation models show no obvious network effect or sustainable differentiation, the chatbot is a limited v1 interface, and the model labs cannot build every application, so the value will likely move up the stack the way it did with chips, ISPs, and mobile networks rather than staying with the model providers. He covers the brutal supply and demand disequilibrium driving today’s token pricing and ten thousand dollar surprise bills, the financial gravity problem of hyperscalers spending over half their revenue on capex, the Jevons paradox and consumer surplus that may compete away productivity gains, the way the important questions move out of San Francisco and into industries like law, consulting, finance, and advertising, and the distinction between automating tasks and changing jobs. His closing image is an IBM ad from the 1950s promising “150 extra engineers,” a reminder that every platform shift feels unprecedented and that in twenty years we will simply say of course computers do that.

    Thoughts

    The most useful thing Evans does here is refuse to collapse uncertainty into a clean prediction, and then explain exactly why that refusal is the correct posture rather than a cop out. He distinguishes between the parts where he will commit to a view, that foundation models are probably not a product and the chatbot is probably not the right interface, and the parts where there are simply too many open paths to call. That discipline is rare in AI commentary, where the incentive is to sound certain. The commodity argument is not “models are worthless.” It is a chain of reasoning: there is no visible network effect, no durable differentiation beyond willingness to spend, no lock in comparable to Windows or iOS, and a likely structure of three to six well funded competitors plus open source and edge models all selling the same thing. Ask where price discipline comes from in that picture and the honest answer is that it probably does not, which is how you get a commodity even when demand is effectively infinite.

    The mobile data analogy is the load bearing comparison and it deserves to be taken seriously. Mobile data traffic rose something like fifteen hundred to two thousand times over fifteen years, the networks built an extraordinary piece of global infrastructure, everyone came to depend on it, and yet the operators captured almost none of the value because all the interesting stuff got built on top by someone else. Telco stocks were flat for two decades. If that is the template, then the trillion dollars of capex flowing into AI infrastructure can be both a worthwhile investment and a terrible place to expect outsized equity returns, because building the road is not the same as owning the traffic. The counterpoint Evans keeps fairly on the table is the operating system path, where Windows and iOS did capture value, but he notes they had levers and network effects that LLMs do not appear to have.

    His framing of where the questions live is the part most people in tech underweight. Once a technology works, the interesting questions stop being technology questions. Netflix is not a tech company in the sense that matters, because its real decisions are Los Angeles decisions about shows, talent, and sports, not San Francisco decisions about infrastructure. By the same logic, what AI means for a law firm is mostly a question for people who understand what associates actually do and what clients are actually paying for, not for model researchers. This is why the “the model will just do the whole thing” story keeps running aground. Most valuable software does not solve a problem the customer already knew they had. It often takes years to convince an industry that a problem even exists, and an LLM prompt does not surface latent problems that no one has articulated.

    The economic plumbing he describes is where the near term risk actually sits. We are in extreme disequilibrium, where twenty dollars a month can buy ten thousand dollars of tokens on one side and a weekend of experimentation can produce a ten thousand dollar bill on the other, exactly the pattern mobile data went through around 2009 and 2010. That gets resolved with the boring machinery of caps, throttling, and pricing tiers, not with magic. Layered on top is the financial gravity problem: Microsoft, Meta, and Google heading toward spending more than half of revenue on capex, with roughly seven hundred billion dollars of guidance across the big players, against a hard ceiling because there is not ten trillion dollars a year available to spend. And even when the productivity gains are real, the Jevons paradox and consumer surplus suggest much of the benefit gets competed away. If a discounted cash flow model used to take a week and now takes ten seconds, you do fifty of them and charge the client the same, which is great for clients and unremarkable for margins.

    The honest takeaway for builders is that the answer to “what does this do to software” is more software, probably one or two orders of magnitude more, just as SaaS itself produced an explosion rather than a consolidation. The SaaS apocalypse is real in the sense that some meaningful percentage of existing companies get wiped out, and unknowable in the sense that no one can yet say which ones, which is why thoughtful investors are reluctant to be long software in the dark. For anyone pursuing a more deliberate, purposeful relationship with technology, the closing note is the one to keep: every one of these shifts felt singular and world ending and world making at the time, it reshaped work and put people out of jobs and created things we love, and then it quietly became invisible. The goal is to stay clear eyed about which of those buckets a given change lands in rather than getting swept up in the noise of what someone said at a party yesterday.

    Key Takeaways

    • Agentic coding shifted from “kind of useful” to “really changing everything” at the start of the year, and it is the single field with unambiguous product market fit, where customers are pulling it out of your hands.
    • Coding working first was foreseeable in hindsight: software developers were the ones messing with the tools, and the first thing people do with a new kind of computer is build more computing, just as the first thing people did with PCs was make computers.
    • Anthropic, with less capital raised, chose to focus on coding and got it working, while OpenAI cycled through a more everything all at once strategy before narrowing in.
    • The intense focus on coding comes bundled with a supply crunch, a capacity crunch, and a price and capex imbalance that defines the current moment.
    • Most of the fundamental questions from two or three years ago still have no answers: whether there will be a winner in models, whether models capture value up the stack, how much they can do, and whether consumers will use this daily rather than weekly.
    • There is a wide gap between Valley insiders running clusters of Mac Studios all day and the roughly forty percent of people who say AI is “kind of useful, I used it last week for something.”
    • Outside tech, companies are adopting AI as one at a time point solutions for specific back office processes, like a commodities company using LLMs for better cash flow forecasting, not as a general purpose assistant.
    • Adoption always compounds on prior platforms: you could not have nine hundred million weekly active users in the Netscape era because there were not nine hundred million PCs on the planet.
    • Early in any platform shift almost nothing works smoothly, from sound cards and floppy disks with TCP/IP to computers that froze and lost your work, and AI is at that stage now.
    • Today’s token pricing crunch mirrors the mobile data shock of 2009 to 2010, where flat rate plans collided with surging usage and networks had to realign price with marginal cost through caps, fair use, and throttling.
    • Mobile data traffic rose roughly fifteen hundred to two thousand times in fifteen years, mobile networks earn around a trillion dollars and spend about two hundred billion a year on capex, yet their stocks have been flat for twenty years because all the value moved up the stack.
    • The central LLM question is whether the model can do the whole thing or whether you need hundreds of applications built on top, the same way you needed apps on Windows and iOS.
    • Evans sees no network effect and no sustainable differentiation between models beyond willingness to spend money, which points toward commodity infrastructure sold near marginal cost.
    • Chip companies, ISPs, and mobile operators did not capture the value; Windows and iOS did, but only because they had levers to move up the stack and real network effects, which models lack.
    • A useful comparison is semiconductors, where each generation gets more expensive and the field narrows to fewer players, suggesting three to six frontier model makers spending somewhere between two hundred billion and two trillion dollars a year.
    • Enterprises do not standardize on a model the way they once thought about AWS; the cloud and the model get abstracted away, so customers do not even know which one their SaaS product runs on.
    • Demand for tokens being effectively infinite does not prevent a price equilibrium, exactly as infinite demand for mobile bits still produced murderous price wars between commodity carriers.
    • History teaches that something will happen but rarely what; the smartest people in tech wrongly predicted Android would crush the iPhone on open versus closed grounds.
    • One characteristic of tech is that the moment you understand how something works is the moment to move on, which is why Evans stopped updating his Apple spreadsheet years ago.
    • The people who are good at using a tool are usually not the people who are good at designing what the tool should be, which is why model labs cannot build every skill or vertical application.
    • Claude skills and similar templates resemble file new in Excel: useful starting points that users eventually outgrow, raising the question of who builds the real software.
    • The questions increasingly move out of technology and into specific industries; what AI means for law, consulting, advertising, or accounting is partly an AI question and partly a deep domain question.
    • Netflix is not a tech company in the way that matters, because its real questions are media industry questions about shows, talent, and sports, not infrastructure; the same logic now applies across industries facing AI.
    • AI differs from prior platform shifts because the physical limits are unknown; in 1995 you knew PCs cost three thousand dollars and broadband could not reach everyone overnight, but no one knows how cheap, fast, or capable models will get.
    • Evans offers four buttons to press on any use case: is it just price elasticity and the Jevons paradox, does it remove a cost barrier to entry, does it unlock a new business model, or does it make something previously impossible now possible like trains over horses or Spotify over CDs.
    • Advertising and e-commerce are a standout opportunity because today’s systems know a SKU and a metadata field but not what a product actually is or why people buy it, and LLMs could change that level of understanding.
    • The valuable shift is not doing the old thing more, like more spreadsheets or better email, but doing genuinely new things, such as asking an LLM how to change prices to improve churn using all your call recordings, CRM flows, and product telemetry.
    • Enterprise software today splits into three buckets: big horizontal systems like SAP and Workday, three to four hundred vertical SaaS apps plus a thousand internal apps, and a fuzzy improvised middle of Excel, email, and shared files, with AI arriving as a new option across all three.
    • A core design tension is where to put the probabilistic software that can make mistakes versus the deterministic database that cannot, and whether the LLM sits at the top or the bottom of the stack; the answer is probably both depending on the task.
    • The net effect on software is way more software, since SaaS itself produced one to two orders of magnitude more software and all software companies exist to solve problems created by other software companies.
    • The SaaS apocalypse is real but unknowable: some percentage of SaaS companies get wiped out, but no one knows which, so you should not derate the whole sector fifty percent and many investors are wary of being long software for now.
    • Much of what an organization does is implicit, undocumented, and not in the training data, which is exactly the value McKinsey, Bain, and BCG provide by getting license to map how a company really works.
    • The real decisions are usually exception handling: the question is always what you cannot automate and what still requires human judgment about cases that were never written down.
    • Distinguish tasks from jobs: accountants spend almost none of their time the way they did fifty years ago, yet to the client the job looks the same.
    • LLMs excel where you want the average, the answer anyone would give, and struggle where you specifically do not want the average and cannot fully explain why you did it differently.
    • There is a financial gravity ceiling: Microsoft, Meta, and Google are on track to spend over fifty percent of revenue on capex versus fifteen to twenty percent for capital intensive telecoms, with seven hundred billion in guidance this year and no path to ten trillion.
    • Hyperscalers face an existential FOMO trap: returns look positive now, but they cannot let rivals build the future of compute without participating, even as the CFO asks how much participation is enough.
    • Token maxing will face a reckoning as the disequilibrium resolves, but measuring ROI is hard because most reported benefits so far, like better analytics, support, and productivity, are tough to put a financial value on.
    • Consumer surplus means many gains get competed away: if analysis that took a week now takes a day, you do five times more analysis and charge the same, the way investment banks did with spreadsheets.
    • Evans closes with a 1950s IBM ad promising “150 extra engineers,” a reminder that every fundamental technology change feels unprecedented, and that in twenty years AI will simply be invisible magic we take for granted.

    Detailed Summary

    What changed in the last year

    Evans frames the past year as a narrowing of focus. A year and a half after the first version of his presentation, the field has developed a much clearer sense of diverging product strategies and competitive tension that goes beyond simply building a bigger model with more compute. The dominant shift is that agentic coding started genuinely working, and the entire industry narrowed in on it because it has absolute product market fit, the kind where customers pull the product out of your hands. That success arrives alongside the supply crunch, capacity constraints, and price imbalance that now define the moment. At the same time, the charts keep climbing, models keep getting bigger, capex keeps growing, and usage keeps growing, while the deep questions from a few years ago remain unanswered.

    Why coding worked first

    That coding led was predictable at a naive level: the people experimenting with the tools were software developers, and they naturally tried to make software development work. Evans compares the moment to the internet around 1997 and 1998, and also to PCs in the late seventies and early eighties, when the technology was exciting but it was not clear what it was for and it did not quite work yet. The first thing people did with PCs was make computers, and since LLMs are in a sense computers, the first thing people are doing with them is making more compute. What was harder to foresee was the precise timing of the shift, the moment when agentic coding flipped from useful to transformative at the start of this year.

    Jobs, juniors, and what we have not learned

    On the question of what this means for engineers and team structure, Evans is blunt that we have learned almost nothing yet, because this did not even work six months ago and everyone is scrambling to interpret it. The pricing crunch alone means it will take a couple of years to settle. The newly concrete questions include whether you still hire junior people and what they would do, and why you were hiring juniors in the first place, whether to do the work itself or to develop people. Because software development now genuinely automates a class of work that used to be done by people, those questions have moved from theoretical to real, but no one can responsibly claim to know what a software team or a software career looks like in three years.

    OpenAI, Anthropic, and the strategy split

    Evans dryly notes the drama around the model labs, including the disruption of a senior leadership medical leave at OpenAI. In the latter part of last year, OpenAI’s question was essentially what to build on top of the models, an everything all at once approach that looked almost like asking the model for fifteen ideas and then doing all of them. Anthropic, with less capital raised, instead committed to coding and got it working, whether by deliberate strategy or by stumbling into it. The result is that software development plus a few other fields are where things genuinely work, surrounded by a large population of people excited around the edges and corporations quietly automating specific back office processes. He cites a commodities company that wants LLMs for better cash flow forecasting across many small producers, a very different thing from asking a chatbot to summarize your meetings.

    The mobile data analogy and value capture

    The richest section is the comparison to mobile. Adoption always compounds on prior platforms, so AI inherits a far larger installed base than the internet or mobile did at their starts. Early on, nothing works smoothly, and Evans recalls the era of buying a three hundred dollar sound card or wrestling a floppy disk of TCP/IP into a machine. The pricing dynamics directly echo mobile data around 2009 and 2010, when flat rate plans met exploding usage and ten thousand dollar bills, forcing networks to realign price with marginal cost. Crucially, mobile data traffic then rose fifteen hundred to two thousand times, the networks built extraordinary global infrastructure with around a trillion dollars of revenue and two hundred billion in annual capex, and yet their stocks stayed flat for twenty years because all the cool stuff and all the value got built and captured by someone else higher up the stack. Chip companies, ISPs, and mobile operators did not capture value; Windows and iOS did, but they had levers and network effects that models do not appear to share.

    The case that models become commodities

    Evans lays out the building blocks of his commodity thesis. First, there is no clear way to build a model that is sustainably and fundamentally better than everyone else’s, with no visible network effect and no strategic lever comparable to what Instagram, YouTube, or Google search enjoy. Differences in emphasis and taste exist, but not durable competitive moats beyond spending. Second, the chatbot is a weird, limited v1 interface that works well for some tasks and people but requires tooling, the right data, configuration, control, and thoughtful design for most real jobs, and the people good at a job are rarely the people good at designing the tool for it. Third, the labs cannot build every application any more than Microsoft or Apple could build every Windows or iPhone app. Enterprises do not standardize on a model the way they never standardized on a visible cloud provider, because it gets abstracted away. Taken together, that points to low level infrastructure sold by perhaps half a dozen competitors plus open source and edge, with no obvious source of price discipline, which is the definition of a commodity even when demand is infinite.

    The questions move out of technology

    One of the next big questions is when models become good enough that you no longer need the largest, fastest, most expensive model, and can use an older model, an open source model, or one running on device where compute is effectively free to the developer. But the deeper shift is that the important questions move out of technology and into industries. Drawing on his own essays “content isn’t king” and “Netflix isn’t a tech company,” Evans argues that Netflix’s real decisions are Los Angeles media questions, not San Francisco infrastructure questions, and San Francisco does not even know what the right questions are. By the same logic, what AI means for a law firm is mostly a question for people who understand law firms, what generative video means for Hollywood is a question Ben Affleck can answer better than he can, and the questions become half AI and half something else.

    Four buttons and the new things AI unlocks

    To reason about impact, Evans offers four buttons. Is a use case just price elasticity, the Jevons paradox of doing the same thing for less or more for the same money. Does it remove a cost that was a barrier to entry, like a newspaper’s printing press. Does it unlock something in your business model. Or does it make something previously impossible now possible, the way steam engines made trains possible regardless of how many horses you bought, or Spotify turned fifteen dollars a month into all the music there is. He stresses that the same broad change can mean wildly different things by industry, just as the internet devastated newspapers but barely touched movie studios. His favorite tractable example is advertising and e-commerce, a trillion dollar advertising market against twenty five trillion in retail, where today’s systems know a SKU and a metadata field and that people who bought one thing bought another, but do not know what a product is or why people buy it. An LLM could in principle understand the product, recommend ten coats at different prices with pros and cons, or look at your Instagram and suggest a winter coat that changes your look but not too much, which would have been science fiction three years ago.

    More software, the SaaS apocalypse, and tasks versus jobs

    For software specifically, Evans expects more competition, cheaper and quicker building, and new categories that were impossible before, all under an uncertain new margin structure where outcome based pricing is hard because most software work cannot be tied cleanly to profit and loss. He frames enterprise software as three buckets, big horizontal systems, hundreds of vertical and internal apps, and a fuzzy improvised middle of Excel and email, with AI arriving as another option across all of them. The deeper design tension is where to place probabilistic software that can make mistakes versus deterministic systems that cannot, and whether the LLM sits at the top or bottom of the stack, with the answer being both depending on the task. The net result is way more software, since SaaS itself produced orders of magnitude more software and software exists to solve problems created by other software. That fuels the SaaS apocalypse anxiety: some companies clearly get wiped out, but since no one knows which, you should not derate the whole sector, even as many investors stay cautious about being long software.

    Implicit knowledge, exception handling, and where the average fails

    Much of what organizations do is implicit, undocumented, and absent from any training data, which is precisely the value of strategy consultancies that get license to map how a company really works versus how it is supposed to work. The real decisions tend to be exception handling, the cases that require human judgment because they were never written down or do not look like before. Evans separates tasks from jobs, noting accountants do almost nothing the way they did fifty years ago while the client still buys the same thing. And he offers a sharp test: LLMs are excellent where you want the average, the answer anyone would give, and weak where you specifically do not want the average and cannot fully articulate why you did it differently.

    Capex, financial gravity, and the ROI question

    On spending, Evans describes a financial gravity problem. Microsoft, Meta, and Google are on line to spend over half their revenue on capex this year, against fifteen to twenty percent for capital intensive telecoms, with roughly seven hundred billion in guidance across the big players, a sum comparable to all of telecom or oil and gas. They cannot sustainably leap to one and a half trillion next year because the money is not there, so the curve must eventually taper. The hyperscalers are caught in an existential FOMO trap: returns look positive now, but they cannot sit out what might be the future of compute without risking becoming the next stranded incumbent, even as the CFO asks how much is enough. On token maxing, he expects a reckoning as the disequilibrium resolves, but measuring ROI is genuinely hard because most reported benefits so far are soft and hard to value, and consumer surplus means much of the gain gets competed away, the way faster spreadsheets simply meant more analysis at the same price.

    Closing image

    Evans ends with an IBM advertisement from the early 1950s showing a sea of engineers holding slide rules, with the tagline that an IBM electronic calculator gives you 150 extra engineers, exactly the pitch behind countless modern startup decks. We move through these fundamental technology waves every ten or fifteen or twenty years, each one feeling completely unlike anything before, and AI is amazing and transformative in the same way mobile, the internet, and PCs were. The base case is that it will produce wonderful things, ruin some livelihoods, put people out of work, and eventually become invisible. His one line description of where it all ends up is that it will be magic, and in twenty years we will simply say of course computers do that, the way an hour of crash free streaming HD video over Wi-Fi already feels unremarkable.

    Notable Quotes

    “Agentic coding went from being kind of useful to really changing everything.”

    Benedict Evans, on the pivotal shift at the start of the year

    “We are in this extreme scarcity. We can’t spend $10 trillion a year on AI infrastructure cuz there isn’t $10 trillion a year there to spend on it.”

    Benedict Evans, on the hard ceiling of AI capex

    “I don’t think foundation models are a product. I don’t think a chatbot is a product. I think the value will be further up.”

    Benedict Evans, stating the core of his thesis

    “They built this amazing piece of global incredibly sophisticated very expensive global infrastructure with enormous growth in use, and they didn’t make any money from it because all the value moved up stack.”

    Benedict Evans, on the mobile network analogy

    “The moment that you understand something and you know how it works and what’s going to happen is the moment you should move on to something else.”

    Benedict Evans, on how to pay attention in tech

    “These are all Los Angeles questions. These are not San Francisco questions. No one in San Francisco even knows what the right questions are.”

    Benedict Evans, on why Netflix is not a tech company

    “The important stuff is not doing the old thing but more. It’s doing something new that you couldn’t have done with the old thing.”

    Benedict Evans, on where the real value of a new technology shows up

    “All software companies exist to solve problems created by other software companies.”

    Benedict Evans, on why AI produces more software, not less

    “It’s going to be magic, and in 20 years time we’ll just say, well, of course that’s how it is. Computers have always done that.”

    Benedict Evans, on how the whole shift ends up

    This is a dense, clear eyed conversation that rewards a full listen, especially if you are trying to think past the hype cycle about where AI value actually lands. Watch the full conversation here, and check out the “AI eats the world” presentation referenced throughout.

    Related Reading

    • Benedict Evans’ website home of the “AI eats the world” presentation and his newsletter referenced throughout the conversation.
    • Andreessen Horowitz (a16z) the venture firm whose podcast hosted this discussion and where Evans was formerly a partner.
    • Jevons paradox (Wikipedia) background on the price elasticity idea Evans uses to explain how cheaper AI may lead to more usage rather than savings.
    • Stratechery by Ben Thompson the analysis Evans cites on software as a designed workflow versus a process that grows out of how a business runs.
    • The Pursuit of Purpose a PJFP look at finding direction and meaning in work as automation reshapes careers and industries.

  • Whale Rock Capital Founder Alex Sacerdote on S-Curve Investing, Why Anthropic Is His Highest Conviction Bet, and the Decommoditization of AI Hardware

    Alex Sacerdote built Whale Rock Capital into one of the most respected technology hedge funds in the world by treating markets through a single disciplined lens: the technology adoption S-curve. In this long conversation on Invest Like the Best with Patrick O’Shaughnessy, he lays out the full framework that has carried him through internet 1.0, mobile, cloud, e-commerce, and now AI, and he explains why Anthropic became his highest conviction position, why his fund went net short application software, and why the least glamorous corner of the market, the hardware and chips that build out data centers, may be one of the best ways to play artificial intelligence right now. What follows is the working theory of a money manager who has spent twenty years trying to think exponentially while the rest of the market thinks one quarter at a time.

    TLDW

    Sacerdote walks through Whale Rock’s three-part investment framework: find the right part of an S-curve, identify the company with a durable competitive advantage, and buy when long-term earnings power is underappreciated. He tells the story of investing in Anthropic at a 180 billion dollar valuation in August 2025 after Claude Code made coding the true unlock of AI, and frames the foundational model market as a three-horse race between Anthropic, OpenAI, and Google that resolved from sixty startups into an oligopoly. He argues enterprise AI is less than 1 percent penetrated, calls the adoption shape an L curve rather than an S-curve, and warns there is not enough compute in the world. He explains why he sold almost all of his application software and went net short, why he loves the decommoditization of AI hardware (Celestica, Corning, Elite Materials, Delta, Advanced Energy, high bandwidth memory, 40-layer PCBs), introduces a modified rule of 40 for chip investing, surveys the moats that let leaders win (network effects, industry standard, scale, critical IP, brand, recursive self-improvement), discusses moving from public markets into private deals like Stripe and Anthropic, lays out Whale Rock’s fund products including the new Mega Cap Tech Fund, defends old-fashioned scuttlebutt research in an AI age, and closes on the kindest thing anyone ever did for him, his father joining the firm after 41 years at Goldman Sachs.

    Thoughts

    The most useful idea in this conversation is not the bullishness on AI, which is everywhere now, but the discipline underneath it. Sacerdote’s framework forces a separation that most investors collapse. A great market is not a great investment. A great company is not a great investment. You need a tall S-curve, a company with a moat that survives the curve, and a price that does not yet reflect the earnings power. He says the quiet part out loud: he has repeatedly bought the best companies in the world at four or five times earnings precisely because the market refuses to extrapolate exponential growth. Nvidia at four times earnings in 2023, Tesla at five times in 2019, Amazon where AWS came free. The edge is not information, it is the willingness to underwrite two to four years out when the consensus cannot see past the next quarter.

    The Anthropic story is the framework applied in real time, and it is worth noting how late and how cautious he was. Whale Rock passed on the 60 billion dollar round because gross margins were negative and coding had not yet exploded. They only got conviction once Claude Code flipped from autocomplete to agentic work, once they heard Anthropic engineers were burning 100 dollars a day in tokens, and once the math on twenty million coders implied a half trillion dollar market from coding alone. The lesson he repeats throughout, that it is okay to be late, that you can miss the first 100 percent if the curve is tall enough, is a direct rebuke to the fear of missing out that drives most AI investing. He waited for the moat to be visible before he paid up.

    His most contrarian and most actionable call is on hardware. The consensus reflex is that chips and components are commodities that get competed to zero. Sacerdote argues the opposite is happening: AI workloads growing 10x a year are pushing every layer of the server to its physical limits, and that pressure is decommoditizing the entire stack. A liquid-cooled AI server is a 300,000 dollar piece of critical infrastructure, not a 5,000 dollar throwaway box, which means the supplier becomes a permanent fixture like a parts vendor on a plane. The Celestica example is the template: a contract manufacturer left for dead since 1999 that turned out to be the sole supplier of Google’s TPU server and a leader in liquid cooling and Ethernet switching, trading at eight times earnings. If he is right that we are 30 percent short on DRAM, NAND, and PCBs, the picks-and-shovels trade has years left to run regardless of which model company wins.

    The software bear case deserves the most scrutiny because it is the most consequential and the least certain. Going from 40 to 50 percent of the portfolio in software to net short is a violent reallocation, and his reasons are layered: AI products that nobody will pay for, CIO budgets being raided to fund Anthropic tokens, pricing power evaporating, and the long-term threat that AI-native startups rebuild incumbents from scratch. But he is honest that the bull case is real too, that old technology is sticky, that companies prefer to buy rather than build, and that AI might actually make platforms like Slack or CRM more important if agents end up operating inside them. This is the genuine uncertainty in the whole AI trade. The bottom of Jensen’s cake, chips and models, is where the value has accrued so far, but historically the application layer captured most of the market cap. Sacerdote is betting that this time the infrastructure and model layers hold the value longer, and he admits the application ecosystem is still unclear and a little bit dangerous. That admission is more valuable than any of his confident calls.

    Finally, the section on research in an AI age is a quiet refutation of the idea that this work automates away. Sacerdote runs a Philip Fisher scuttlebutt operation, 2,500 to 3,000 face-to-face management meetings a year, two decades of compounding relationships, the tripod of conviction where he, his analyst, and a respected outsider all independently like an idea. AI writes better notes now, but the paragraph on top, the wisdom about what it means and how it fits the thesis, is still human. The durable moat in his own business is the same one he looks for in the companies he buys: an accumulated advantage that newcomers cannot replicate quickly. That consistency between how he invests and how he operates is the most credible thing in the interview.

    Key Takeaways

    • Whale Rock’s framework has three legs: identify the right part of a technology S-curve, find the company with a powerful competitive advantage, and invest when long-term earnings power is underappreciated.
    • The core insight is exponential, not linear. Strong tech business models grow earnings exponentially, and because the market refuses to extrapolate, you can buy elite companies at very low multiples.
    • Concrete examples of buying exponential growth cheaply: Nvidia at four times earnings in 2023, Tesla at five times in 2019, Apple at four times, and Amazon where AWS was effectively free.
    • When ChatGPT launched in November 2022, Whale Rock did a firm-wide deep dive and chose to invest in chips and infrastructure first, because demand arrives there first and the winners are knowable regardless of who wins the model layer.
    • The foundational model market went from roughly 60 startups to a three-horse race: Anthropic, OpenAI, and Google. Most startups died, Amazon never showed up, and Meta faltered and had to reboot.
    • Anthropic was the dark horse that focused purely on enterprise while OpenAI won consumer. Whale Rock made it their highest conviction position.
    • Coding is the true unlock of AI. The progression went from Microsoft Copilot at 20 dollars a month (fixing grammar, finding a bug) to Claude running agentically and writing most of the code.
    • The market math: Anthropic engineers were reportedly spending 100 dollars a day on tokens, roughly 20 to 30 thousand dollars a year, and with about 20 million coders in the world that implies a half trillion dollar market from coding alone.
    • Whale Rock invested in Anthropic at the 180 billion dollar valuation in August 2025, when the company hoped to reach 9 billion in revenue and nobody yet knew what 2026 could be.
    • Andrej Karpathy and Linus Torvalds both flipped on AI coding. Karpathy went from 80 percent handwritten code to writing almost no code except in English.
    • Models are not pure commodities. There is real differentiation: Anthropic is strong for private equity and finance, Google is strong at ingesting PDFs, and routers that switch between models mask but do not erase that differentiation.
    • Anthropic is building an ecosystem around the API (SDK, orchestration, the harness, tools), echoing how AWS built lock-in with products around commodity servers starting in 2013.
    • The 800 million people using AI are mostly using AI 1.0, a search engine on steroids. Sundar Pichai estimated only about 10 basis points of knowledge workers are truly using AI’s new capabilities.
    • Enterprise AI is less than 1 percent penetrated. Whale Rock calls the adoption shape an L curve or backwards L curve because it goes straight up, unlike the slower 30 to 50 percent growth of cloud and SaaS.
    • There is not enough compute in the world. Anthropic reportedly has half of what it needs, and Marc Andreessen said the one thing he is sure of is that there will not be enough compute for the next four years.
    • The infrastructure S-curve is only about 10 percent penetrated and remains one of the best ways to play AI.
    • Getting into private deals requires a double opt-in. Whale Rock did a 90-page deck (built with Claude Code) on the coding market to win their Anthropic allocation, and their first private was Stripe in 2020 at a 35 billion dollar valuation.
    • The unicorn private market is now bigger than most European stock markets, larger than Germany or the UK individually. Whale Rock does 2,500 to 3,000 management meetings a year, 10 to 15 percent with privates.
    • S-curves come in two sizes: mega S-curves (internet, mobile, cloud, e-commerce, AI) and sub S-curves within them. AI is the biggest of all and each curve builds on the last.
    • Adoption inflects when barriers fall. Steve Jobs cut the smartphone price to 200 dollars on a 3G touchscreen, Elon cut the EV price to 40,000 with 300-mile range and a working supply chain. Remove the barriers and you get the tornado of demand.
    • Knowing how tall the curve is tells you when to sell. Growth stops being exponential around 30 to 40 percent penetration, when the sell side catches up and big beats end. EVs hit a wall at 10 to 15 percent instead of the expected 40 to 50 percent.
    • Selling Apple in 2012 at roughly 50 percent US smartphone penetration was a mistake, because the moat let it keep compounding around 20 percent even after the explosive phase ended.
    • At strategic inflection points you cannot trust the data (Andy Grove). The signal is intuition and anecdote: a 12-year-old in China on a giant phone playing a real game, or standing-room-only sessions at the Gartner IT Symposium for AWS, VMware, and Splunk.
    • Adoption slope varies. The radio curve hit near-full penetration in about 7 years, while B2B and infrastructure (the dishwasher that has to be plugged in) take far longer. AI is fast because you just open a browser.
    • The moats that let leaders win: network effects, becoming an industry standard, rapid scale, critical intellectual property, brand, and platform lock-in. Anthropic appears to have critical IP, enterprise brand, escape velocity, and recursive self-improvement from using its own code on its own models.
    • On the internet, the leader usually goes bigger, faster, and wins, and compounds on itself (Amazon, Shopify). Exceptions come at paradigm shifts, like AOL failing to make the dialup-to-broadband transition.
    • Whale Rock went from 40 to 50 percent in software five years ago to net short entering this year, which helped performance in the first quarter. AI products were not good enough to charge for and were not moving the needle.
    • Software faces a stack of headaches: falling priority on CIO to-do lists, budget pressure from token spend, lost pricing power, hiring freezes that hurt seat-based models, and the long-term threat of AI-native replacements.
    • The classic rule of 40 is growth rate plus operating margin. Whale Rock’s modified rule of 40 for chip investing is percent of sales that are AI plus market share in that category. Software AI exposure is still only 1 to 2 percent.
    • AI may make some platforms more important. The first thing you do with Claude is plug it into Slack, which could make Slack a permanent repository, and agents may end up operating inside incumbent tools like CRM, solidifying rather than killing them.
    • The data center stood still for 40 years on Intel x86, with every component commoditized. AI changed that. Workloads growing 10x a year are driving the decommoditization of the hardware industry.
    • Celestica is the template: a contract manufacturer left for dead since 1999, sole supplier of the Google TPU server, strong in liquid cooling and Ethernet white-box switching, with 50 to 60 percent share of the cloud Ethernet switch market, once trading at eight times earnings.
    • The whole supply chain is rerating: high bandwidth memory stacked 10 chips high, 40-layer PCBs (versus 10 for a normal server), Elite Materials copper clad laminate, Corning fiber (enough to circle the world four and a half times in one Microsoft data center), and Delta and Advanced Energy power supplies seeing ASPs rise 40 percent a year.
    • Networking has three layers: scale out (racks together), scale across (data centers together), and scale up (every GPU in a rack, currently copper, eventually fiber). The copper-to-fiber shift could two-to-three-x Corning’s opportunity.
    • Whale Rock estimates the market is roughly 30 percent short on DRAM, NAND, and PCBs even at today’s 10 basis points of real AI usage.
    • Rate of change matters more than absolute level. When Claude plotted market share data it missed the rate of change, the thing that drives accelerating growth and margins as a company moves from 10 to 30 percent share.
    • Key risks: public and government negativity toward AI (Maine reportedly banned data centers, only 20 percent of people are optimistic), models hitting a wall and letting open source catch up into a race to the bottom, and a major player faltering and stranding compute.
    • Chip companies do not care who wins the token war, which makes them a relatively safe way to play AI. Jensen Huang actively wants open source to take off.
    • Research is still human work. Whale Rock runs a Philip Fisher scuttlebutt process, the tripod of conviction (Alex, the analyst, and a respected outsider), and 20 years of compounding knowledge. AI writes better notes but cannot supply the wisdom paragraph on top or pick stocks.
    • The firm’s product evolution: 15 years as a long short fund, a long only fund in 2020 that is now larger than the long short, opt-in privates formalized around 2015 and activated in 2020, an 80 percent privates hybrid fund in 2021, and the new Whale Rock Mega Cap Tech Fund.
    • The Mega Cap Tech Fund thesis: endowments are structurally underweight the largest tech companies because they believe there is no alpha in large cap. Whale Rock takes the top 30 global market caps and picks the best 12 or 13, arguing it takes 100 diversified PMs to realize Google is a winner.
    • The kindest thing anyone ever did for Sacerdote: his father, after 41 years at Goldman Sachs, joined Whale Rock as chairman and the gray hair for six years until he passed away in 2011.

    Detailed Summary

    The Anthropic Investment and the Three-Horse Race

    When ChatGPT launched in November 2022, Whale Rock immediately took its 10-person team and ran a firm-wide deep dive. Sacerdote’s first principle is that every new compute paradigm creates a new stack with new winners and losers, and in this stack the layers run from power and chips at the bottom, to the clouds, to the foundational models, to the applications on top. In early 2023 the firm deliberately positioned in chips and infrastructure first, reasoning that demand arrives there first and the winners are knowable no matter who wins above. At an April 2023 webinar they framed the model layer as a coin flip between winner-take-all, total commodity, a race to zero, or an oligopoly of three or four. Over the next three years the answer became clear: of roughly 60 startups, almost all died, Amazon never really showed up, Meta came in strong then faltered and rebooted, and Anthropic emerged as the dark horse focused purely on enterprise while OpenAI won consumer and Google remained a perennial threat. The result looked like the cloud market, where three companies underpin the entire SaaS world with excellent businesses.

    The decisive factor was code. Sacerdote says the firm was initially skeptical AI could replace labor, given the negative corporate feedback on early models. That changed in 2025 when Claude Code and the agentic coding tools exploded. The progression ran from Microsoft Copilot at 20 dollars a month, which could improve coding grammar or find a bug, to Claude running agentically and doing far more. The token economics were staggering: Anthropic engineers reportedly spending 100 dollars a day, which annualizes to 20 to 30 thousand dollars, and with 20 million coders worldwide that implied a half trillion dollar market from coding alone, on technology that was only 7 to 9 months old. Whale Rock made the investment at the 180 billion dollar valuation in August 2025, writing in their letter that the company hoped to reach 9 billion in revenue, with growth like nothing they had ever seen, 100 million to a billion on the way to 9 billion, and no one yet knowing what 2026 could bring.

    Why the Models Are Not Commodities

    Everyone expected the foundational models to be pure commodities, but Sacerdote argues there is tremendous differentiation within them. Different training methods produce different skills: Anthropic excels at anything touching private equity and finance, Google is strong at ingesting PDFs. Routers that switch between models make them look like commodities but mask genuine, critical IP. Beyond the model itself, Anthropic is building a whole ecosystem around the API: the SDK, the orchestration layer, the tools, and the harness, the software wrapped around the API that gets the most out of the model. He compares this directly to AWS in 2013, when people dismissed cloud as commodity servers in a warehouse and missed that Amazon was inventing products that slowly built lock-in. The open-source risk from China is real, but Sacerdote got comfortable that leading-edge token quality is superior, because going from 80 to 85 percent of benchmark performance is a huge unlock and the open-source players lack the compute to leapfrog the frontier.

    The S-Curve Framework in Full

    Whale Rock’s whole edge is thinking exponentially when the world thinks linearly. Sacerdote argues very few people believe you can accurately predict two, three, or four years out, but if you understand the S-curve, the moats, and how to model, you can. Every technology follows the same pattern: it exists hidden for years (smartphones 10 years before the iPhone, the internet 20 years before Netscape, EVs 15 years before Tesla went vertical in 2019) until the barriers to adoption fall and demand inflects into a tornado. Knowing how tall the curve is tells you when to sell, because exponential growth stops around 30 to 40 percent penetration when the sell side catches up. Curves can also be dynamic: AWS turned out to address a far larger TAM than expected once it became clear cloud was not actually deflationary. There are mega S-curves (internet, mobile, cloud, e-commerce, AI) and sub S-curves within them. AI is the biggest. And slope varies enormously by the nature of the technology, the radio curve hitting full penetration in 7 years, B2B and infrastructure taking decades because, like a dishwasher, they have to be plugged into existing systems.

    On timing, Sacerdote is relaxed about being late. Citing Peter Lynch, who mentored him at Fidelity and told him to white out the chart because it is all about the future, he argues it is fine to miss the first one, two, or three years and even the first 100 percent if the top of the curve is half a trillion. At strategic inflection points, per Andy Grove, you cannot trust the data, so the firm relies on intuition and anecdote: a 12-year-old in China playing a real video game on a huge phone, or the AWS session at the Gartner IT Symposium that was standing-room-only at 9, 10, and 11 in the morning. Spotting the leader pulling away matters because, on the internet, the leader usually goes bigger, faster, and wins, compounding on itself, with exceptions only at paradigm shifts like AOL missing the move from dialup to broadband.

    The Software Bear Case

    Five years ago Whale Rock had 40 to 50 percent of its portfolio in software. Their April 2023 thesis was that incumbents with huge sales forces and proprietary data would take the AI APIs and build great products. Instead, the AI products were not good enough to charge for and did not move the needle, so the firm sold almost all of its application software and entered this year net short, which helped in the first quarter. The bear case is layered: software has fallen down the CIO priority list, budgets are being raided to fund Anthropic tokens with faster ROI, annual price increases look risky, and hiring freezes hurt seat-based models. The deeper threat is that AI-native startups could rebuild any incumbent from scratch, obviating the data advantage. The bull case is genuine too: old tech is sticky (mobile games did not kill consoles, tablets did not kill the PC), companies prefer to buy rather than build, and an ERP is hard to replace. Sacerdote also floats an optimistic twist, that AI could make platforms like Slack more important as agent repositories, and that agents operating inside CRM could solidify rather than destroy it, even as the bear case is that CRM goes headless and gets relegated to a database.

    The Decommoditization of AI Hardware

    This is Sacerdote’s most differentiated call. For 40 years nothing changed in the data center; Intel x86 became the standard, compute grew 25 to 40 percent a year in line with Moore’s law, and every component, from the printed circuit board to memory to enclosures to networking, commoditized. AI broke that. Workloads now grow 10x a year and push every aspect of the hardware to its physical limits, creating both tremendous unit growth and what Whale Rock calls the decommoditization of the hardware industry. He cites Sean Maguire wishing he could run a hardware hedge fund because all the companies are public with powerful IP, and compares it to Sequoia’s best early hardware investments in Apple and Cisco. The economics flip because an AI server is a liquid-cooled, 200 to 300 thousand dollar piece of critical infrastructure where a single failure brings the whole thing down, so suppliers become permanent like a critical part on a plane.

    Celestica is the marquee example: a contract manufacturer that had been a disaster industry since 1999 and went offshore to China, but kept its IBM supercomputing heritage and talent, became the sole supplier of the Google TPU server, and was trading at eight times earnings three years ago. It turned out to be excellent at liquid cooling where others failed, holds 50 to 60 percent share of the crucial cloud Ethernet switch market, and its engineers helped write the open-source SONiC software, working closely with Broadcom. The same dynamic runs up and down the chain: high bandwidth memory stacked 10 chips high that took Samsung years to master, 40-layer PCBs versus 10 for a normal server with very few suppliers able to make them, Elite Materials supplying the copper clad laminate, and Corning’s fiber, thinner and more bendable, with enough in a single Microsoft data center to circle the world four and a half times. Networking splits into scale out, scale across, and scale up, with the eventual copper-to-fiber shift in scale up potentially two-to-three-x-ing Corning’s opportunity. Power supplies from Delta and Advanced Energy are seeing ASPs rise 40 percent a year at higher margins because each Nvidia rack uses 50 to 125 percent more power. Visibility has gone from we’ll call you next week to design this roadmap with us for four years, turning 5 percent low-margin businesses into 35 to 50 percent topline growers with rising margins, and the whole market is roughly 30 percent short on DRAM, NAND, and PCBs.

    Private Markets, Risks, and the Research Machine

    Moving from public markets into privates meant adapting to a double opt-in, where the company has to choose to let you in. Whale Rock won its Anthropic allocation partly by building a 90-page deck with Claude Code scouring the internet for feedback on the coding market. Their first private was Stripe in April 2020 at a 35 billion dollar valuation, which they could only underwrite because they knew the public comp Adyen cold, and they upsized to a 100 million dollar block. The unicorn market is now bigger than most European stock markets combined. On risk, Sacerdote worries about public and government negativity (Maine reportedly banning data centers, only 20 percent of people optimistic), the possibility that models hit a wall and open source catches up into a race to the bottom, and a major player faltering and stranding compute, though he notes someone else (like Meta stepping into a cancelled Oracle deal) would likely absorb it, and that chip companies benefit regardless of who wins the token war. He explains his caution on the application layer by noting it always comes later, the iPhone took years to spawn its app economy, and the ecosystem is still unclear and a little dangerous, while pointing to Brett Taylor’s Sierra as the kind of company that could prove it out.

    On the research itself, Sacerdote insists AI has not supplanted the analyst. Whale Rock runs the scuttlebutt approach straight out of Philip Fisher’s Common Stocks and Uncommon Profits, doing 2,500 to 3,000 face-to-face management meetings a year and talking to suppliers, customers, and competitors. AI now writes much better notes and gets the team up to speed quickly on complex areas like ABF substrates, but there must be a wisdom paragraph on top, and it cannot pick stocks or replicate the work two analysts did building conviction in AppLovin and a relationship with Adam Foroughi. He calls the firm the Whale Rock learning machine, a group of 10 highly experienced people compounding knowledge for 20 years, with the tripod of conviction (himself, his analyst, and a respected outside investor all liking an idea) as the test. The firm’s products evolved from a 15-year long short fund to a 2020 long only fund now larger than the original, opt-in privates, an 80 percent privates hybrid in 2021, and the new Mega Cap Tech Fund built on the thesis that endowments are structurally underweight the largest tech companies because they wrongly believe large cap has no alpha. He closes on his father, who left Goldman after 41 years to join Whale Rock as chairman and the gray hair until his death in 2011, a mentor remembered by countless people for his humility and grace.

    Notable Quotes

    “When you get the right part of the S-curve, you get exponential unit growth. If you have a very strong business model, your earnings don’t grow linearly, they grow exponentially.”

    Alex Sacerdote, stating the core of the Whale Rock investment framework

    “The world doesn’t think exponentially. Very few people believe you can accurately predict two, three, four years out. But if you follow and understand the S-curve and you know the moats and you know how to model, you really can predict these great things.”

    Alex Sacerdote, on why the market consistently underprices long-term earnings power

    “The enterprise AI or enterprise application AI market is less than 1 percent penetrated, and we’ve never seen, you know, we talk about S-curves, we call this an L curve, just straight up.”

    Alex Sacerdote, on why AI adoption looks different from every prior technology curve

    “We’re at 10 basis points of people really using AI and we’re already sold out. There’s not enough compute in the world. So Anthropic has half of what they need right now, and that’s before this huge takeup.”

    Alex Sacerdote, on the scale of the compute shortage relative to actual adoption

    “It’s okay to be late. It’s okay to miss the first one, two, three years in a lot of cases, because if the top of the S-curve is half a trillion, the growth can go on for a long time. It’s okay to miss the first 100 percent.”

    Alex Sacerdote, on why fear of missing out is the wrong instinct in a tall S-curve

    “The old way of software is like using a pen and paper or a horse and buggy. The new way of software is like a jet engine or frankly like the transporter from Star Trek. It’s so revolutionary it feels like it has to be disruptive.”

    Alex Sacerdote, explaining why Whale Rock went net short application software

    “You become like critical infrastructure, like selling a critical part on a plane. You’ll never get swapped out.”

    Alex Sacerdote, on how liquid-cooled AI servers turned commodity hardware suppliers into permanent fixtures

    “Why do you tell everyone your secret? It’s like why does the casino teach people how to play blackjack? It’s harder. It’s really hard to do.”

    Alex Sacerdote, quoting his mother on why a public framework does not erase the edge

    “He said, you know, I’ve been at Goldman for 41 years. How about I come and join you? I’ll be the gray hair. I’ll be the oversight. I’ll be the chairman. You do what you do.”

    Alex Sacerdote, recalling his father joining Whale Rock, the kindest thing anyone ever did for him

    Watch the full conversation here: Whale Rock Capital Founder on Investing in the Age of Exponential AI.

    Related Reading

  • Paul Graham and Jessica Livingston on Resilience at Y Combinator: Founder Mode, Cockroaches, Sticking to Your North Star, and Why AI and Climate Keep Them Up at Night

    For the very first episode of Disaster Proof, the conversation goes to a garage in Palo Alto to sit down with Paul Graham and Jessica Livingston, the founders of Y Combinator. They have backed thousands of companies, including many now working in the resilience space, and the discussion covers what makes startups durable, why adaptability beats expertise, how Brian Chesky stumbled into founder mode at Airbnb, why the best ideas grow out of a founder’s own life, and the two specific risks (AI and climate change) that Paul says are the only ones he treats as genuinely game over. You can watch the full conversation on YouTube here.

    TLDW

    Paul Graham and Jessica Livingston explain why constant change favors young, flexible founders, and why Y Combinator picks people over ideas precisely so its judgment never goes obsolete. They unpack adaptability as the trait they hunt for in interviews, the “founder mode” story behind Brian Chesky steering Airbnb through COVID, and the 2008 strategy of funding tough, close-to-revenue “cockroaches.” Paul argues a company survives turbulence by sticking to a North Star instead of acting as a weather vane in shifting moral fashions, using the biosphere tree that collapses without wind as his metaphor for resilience. They turn to climate and energy as the next great market, the difficulty of selling into utilities, the Gridware success story, fusion no longer being thirty years away, and the trap of guilt-based business models versus the reliable assumption that users are selfish, greedy, and lazy. The personal-resilience half covers surviving Twitter mobs, Paul’s obsessive essay process, raising kids by indulging curiosity and picking your battles, prepping by living among reasonable people, political polarization, and why AI and climate are the two things that keep them up at night.

    Thoughts

    The most useful idea in this conversation is also the most counterintuitive: a world that feels like it is ending is structurally good for the people least invested in how it used to work. Paul’s point to terrified founders is that change is only a threat if you have sunk costs in the old order. A young founder has been doing the current plan for two weeks, so a step-function shift in the landscape costs them almost nothing to abandon. The incumbents with elaborate machinery and a decade of assumptions are the ones who should be afraid. That reframes resilience away from defense and toward optionality. The resilient party is not the one with the thickest walls, it is the one with the least to unlearn.

    The founder mode discussion is worth sitting with because it quietly overturns a generation of management orthodoxy. The old rule was that a good CEO hires executives and gets out of their way, and that getting into the details is micromanaging. Brian Chesky’s COVID experience at Airbnb broke that rule under maximum pressure. With bankruptcy on the table and a travel company facing a world that stopped traveling, he went line by line through the business and told people what good looked like, then gave them freedom to execute against that standard while still demanding visibility. The interesting nuance is the permission structure. A crisis granted Chesky the license to be involved that normal operating conditions would have framed as meddling. The lesson is not “always be in the weeds,” it is that the founder’s deep understanding and disproportionate caring are assets you are wasting if you reflexively delegate them away.

    Paul’s North Star argument is the part most likely to age well. His claim is that companies fail at resilience when they behave like weather vanes, swinging with each gust of public moral fashion. He pairs it with the biosphere tree that grows weak and topples because it was never exposed to wind. Both metaphors point at the same thing: resilience is built by surviving stress while holding your shape, not by avoiding stress and not by reshaping yourself to whatever the crowd currently rewards. The carbon-credit companies he mentions are the cautionary case. They built their entire premise on a fashion (customer guilt about carbon) and went out of business when the wind changed direction. Durable businesses convert a permanent human motive into value, which is why he prefers the brutally honest assumption that the user is selfish, greedy, and lazy, and that your job is to build something that produces good outcomes anyway.

    The climate and energy section reframes a worthy cause as a market-timing bet rather than a moral appeal, and that is the more powerful version. The comparison to fintech in 2008 is the tell. Banking technology was a sleepy, unglamorous sector that venture investors avoided until a crisis cracked it open and made it one of the best categories of the following decade. The argument is that energy and the physical world are sitting at a similar precipice, made newly viable because hardware is starting to behave more like software (order components, assemble, do not build everything from scratch) and because AI’s hunger for power has made energy the binding constraint on the whole industry. The Gridware story crystallizes the founder lesson underneath all of it. The best founder for a hard physical problem was a lineman who worked the electric lines and lived through the fires. The idea grew authentically out of his life, which is the same pattern Jessica keeps returning to and the same advice they give for raising kids.

    Finally, the personal-resilience material is more practical than it first appears. Paul’s method for surviving a Twitter mob is pattern recognition: once it has happened twenty times, you know it ends in two days and they move on to the next target, so you wait it out instead of capitulating. His essay process is the same conviction-building engine applied to ideas. He goes sentence by sentence until there is no false statement left to attack, which is why his challenge to angry readers (“point out the incorrect statement”) almost never gets answered. The throughline across the company advice, the parenting advice, and the personal advice is identical. You build durable conviction not by sitting in a room thinking, but by working the problem until it is right, then refusing to be blown off course by people who never actually engaged with the substance.

    Key Takeaways

    • Experts are frequently wrong because they are experts in a previous version of the world, so Paul deliberately avoids permanent beliefs about the current state of technology.
    • Y Combinator picks startups by picking founders, not ideas, because the founders know more about the ideas than the investors do.
    • Living in England and visiting for each batch lets Paul arrive every quarter expecting the world to be different, which keeps his mind open instead of anchored.
    • A world of constant change feels bad but is actually good for a young, flexible founder who has only been on the current plan for two weeks and can switch easily.
    • Vibe coding went from kind-of-works to reliably works, and even experienced programmers now generate huge volumes of code with AI.
    • There is still a software business even with AI, because someone has to know what to tell the AI to write, and no company is going to write its own database from scratch.
    • The scenario Paul worries about is model companies spinning up agents to start all the startups themselves, removing the need for human founders.
    • The founder traits Jessica looks for are unchanged over the years: determined, flexible-minded, and willing to adapt.
    • In interviews you can spot rigid founders because they answer the question they prepared rather than the one they were asked, and the gears visibly grind when you redirect them.
    • A good adaptability signal is a founder who says “I haven’t thought about that, but here is how I would think about it” instead of freezing.
    • Founder mode, the term, came from Brian Chesky’s experience steering Airbnb through COVID, when bankruptcy was openly discussed in board meetings.
    • Ken Chenault, the former American Express CEO on Airbnb’s board, told Chesky the moment was ten times worse than 9/11 and could define the company.
    • Founder mode meant Chesky understood every line item, told people what good looked like, then gave them freedom to execute while still wanting to see it.
    • Founders see through the fog because they understand the company better than anyone and they care more than anyone, and combining understanding with caring lets them see more.
    • There is always some disaster at Y Combinator, the way a hospital always has someone coding, so a crisis is the normal operating environment, not an exception.
    • During the 2008 crash, YC kept funding because it is always a good time to start a startup, but focused on people close to making money and very tough founders they called cockroaches.
    • Airbnb was the ultimate cockroach, seemingly indestructible, which is exactly why they liked it during the meltdown.
    • YC rests on two axioms: startups matter, and founders are the most important ingredient in startups. As long as those hold, YC has room to exist.
    • Company values are usually written down a few years in, documenting principles that already existed rather than inventing new ones.
    • You cannot move with fashion; you have to stick to your North Star, especially during turbulent, noisy times.
    • Trees grown inside a biosphere fell over because they were never exposed to wind, so being blown around is a necessary part of becoming strong enough to stand.
    • What preserves YC most is that it is a fundamentally good idea: it gives lonely founders money, the right peers, and colleagues they would never otherwise have.
    • The measure of a good startup idea is revenue, and any other metric you care about matters only because it predicts revenue.
    • At the early stage you can afford to be virtuous and even tell founders to go back to college, because the power law means one startup in the batch will carry the returns.
    • Every startup has to find early adopters, who decide quickly, usually do not have much money, and tend to be sophisticated, which means utilities are rarely your first customer.
    • A company that ultimately sells to utilities should start by selling to something that says yes faster, like running a pilot on a single corporate campus.
    • Utilities are under so much stress from wildfire liability, renewables, EV charging, and AI demand that they are unusually willing to try new things out of necessity.
    • Gridware, founded by a former lineman who lived through major fires, is now backed by Sequoia with PG&E as a huge customer, an example of an idea growing out of the founder’s life.
    • The second-biggest chunk of YC startups after AI is hard tech and physical products, not because software is dead but because building physical things is getting more possible.
    • Energy is one of AI’s fundamental constraints; if Sam Altman could have two things for Christmas, they would be energy and GPUs.
    • Nobody says fusion is thirty years away anymore, and the old thirty-year number existed because it was far enough out to avoid demands for results but close enough to keep attention.
    • Energy and physical markets may be where fintech was in 2008, a sleepy sector about to be cracked open by crisis into a great decade.
    • Guilt is a fragile business model because fashions change what people feel guilty about, which is why carbon-credit companies collapsed when the winds shifted.
    • Assume the user is selfish, greedy, and lazy, then build something that causes good things to happen anyway, like clean power that is simply cheaper and more reliable.
    • To survive Twitter mobs, remember they move on in about two days, half are bots or people you would never talk to in real life, and you cannot become a weather vane for moral fashions.
    • You build conviction by working on and developing an idea, not by sitting in a room thinking, unless it is pure thought like math.
    • Paul writes essays sentence by sentence until nothing in them is false, which is why his challenge to point out an incorrect statement almost never gets answered.
    • The best startup ideas, and the best projects in life generally, grow authentically out of the founder’s own interests and experiences.
    • Their parenting philosophy is to give kids confidence and a stable base, indulge their curiosity, and encourage projects nobody told them to do.
    • You pick your battles with kids: put your foot down on cruelty, but accept defeat on things like food and screen time.
    • A useful interview question for anyone with an unusual experience is not “what was it like” but “how was it different than you expected,” which surfaces the genuinely novel detail.
    • In a time of turbulence, bet on an island full of reasonable people; the English may not be very dynamic, but they are reasonable.
    • The hope on political polarization is to build resilient institutions that act as a cage around any single leader, so that throwing the rattle makes no difference.
    • AI and climate change are the two things Paul worries about most because they are both potentially game over, like the Gulf Stream reversing and turning Europe into a frozen wasteland.

    Detailed Summary

    Staying an expert when the world keeps changing

    The conversation opens on Paul Graham’s essay “How to Be an Expert in a Changing World,” whose core point is that experts are often wrong because they are experts in a previous version of the world. Asked how he keeps his own beliefs from going obsolete when the landscape can shift in ninety days, Paul says he focuses on people. YC picks founders rather than ideas because the founders know the ideas better than any investor could. He deliberately holds no permanent beliefs about the current state of technology, and the rhythm of flying in from England for each batch helps: he arrives every quarter already expecting everything to be different. One quarter the story is everyone training open-source models, the next quarter it is Claude code and nobody bothers with open-source models because the frontier versions are better anyway. He comes in with a completely open mind. Jessica and Paul note that today’s founders are more frightened, asking what is even still true, but the message Paul gives them is that constant change favors the young and flexible. If you have only been executing a plan for two weeks, a disruption costs you nothing; you just switch.

    What adaptability looks like in a founder

    Jessica describes the founders she funds as determined, flexible-minded, and willing to adapt, and calls adaptability a key trait always, but especially in uncertain times. In interviews, the rigid applicants reveal themselves by answering the question they planned to answer rather than the one they were asked, and you can almost hear the gears grind when you redirect them. Paul does not let that slide; if they dodge, he just asks again. The positive signal is a founder who, faced with a question they have not considered, says “here is how I would think about it” and reasons live. Both point out that YC itself had to adapt, and that the company they funded the interviewer’s startup as in 2009 looked very different by the end. They funded him in May 2009, in the thick of the financial crisis, after he had quit his job in August 2008 and briefly felt he had made a terrible mistake.

    Founder mode and seeing through the fog

    Paul points to Brian Chesky as the defining example of weathering disaster, a story he explored on This Week in Startups. When COVID hit a travel company like Airbnb, the word bankruptcy was being used in board meetings, and Ken Chenault, the former American Express CEO on the board, warned it was ten times worse than 9/11. Chesky went into what would later be named founder mode, getting into every line item, understanding exactly what was needed, telling people what good looked like, and then giving them freedom to execute while still insisting on visibility. The crisis gave him permission to be the involved CEO he had always wanted to be, the kind of involvement that normal operating conditions would have labeled micromanaging. Paul argues founders see through fog that blinds everyone else for a simple, rational reason: they understand the company better than anyone because they have been there longest and thought of most of it, and they also care more than anyone. Combine deep understanding with deep caring and of course they see more.

    Cockroaches, the North Star, and the biosphere tree

    Returning to 2008, when YC was self-funded and unsure whether anyone would invest by March, they decided to keep going on the principle that it is always a good time to start a startup, but to fund people close to making money and very tough founders they called cockroaches, after the creatures that survive nuclear war. Airbnb was the ultimate cockroach. Paul frames YC’s longevity around two axioms (startups matter, founders are the most important ingredient) and around resilience built through stress. He tells the story of trees grown inside a biosphere that fell over because they were never exposed to wind, since being blown about is a necessary part of a tree becoming strong enough to support its own weight. YC has been blown around and is still standing, which is exactly what gave it practice. The companion idea is the North Star: you cannot move with fashion or act as a weather vane swinging with other people’s moral fashions, you have to hold your founding principles, which Paul eventually wrote down rather than let a 23-year-old new hire do it.

    Climate, energy, and selling into hard markets

    The interviewer’s own path (a curiosity about wildfire that grew from living in California, watching PG&E go bankrupt, a fire on his Mendocino property, volunteering as a firefighter) becomes the case for ideas that grow authentically out of a founder’s life. Climate is framed broadly as energy, the built environment, and transportation, essentially the physical world, and those are hard markets where the buyers are utilities, governments, real estate, and insurance. The advice is to find early adopters who decide quickly, which usually means not starting with a utility but with something like a single corporate campus that will say yes faster. Utilities, though, are under so much stress from wildfire liability, renewables, EV charging, and AI demand that they are increasingly willing to try new things. Gridware, founded by a former lineman who lived through major fires, is the proof point: backed by Sequoia, with PG&E as a major customer. Paul notes the second-biggest chunk of YC startups after AI is hard tech, not because software died but because building physical things is getting more possible, more like ordering and assembling components. Energy is the binding constraint on AI, fusion no longer feels thirty years away, and the bet is that energy and physical markets are where fintech was in 2008, about to be cracked open.

    Guilt versus greed as a business model

    On the question of whether climate companies should sell on guilt (recycle, pay more because it is sustainable), Paul is blunt that guilt is fragile because fashions change what you are supposed to feel guilty about. The carbon-credit companies thrived until buying carbon credits stopped being cool, then went out of business. A founder’s own concern for the world can drive great companies, but depending on a customer’s guilt is shallow. The durable move is to assume the user is selfish, greedy, and lazy, someone who just wants to eat pizza and watch Netflix, and to build something that produces good outcomes despite that. Clean power is the perfect example: nobody watching Netflix is upset that fusion powers their television, and if it is cheaper and more reliable, that is simply more Netflix and more money for pizza.

    Personal resilience, Twitter mobs, and the essay process

    On surviving public criticism, Paul’s method is pattern recognition: after twenty mobs you stop counting and know it will be over in two days when they move to the next topic, so you wait it out even though it genuinely feels miserable. Half of them are bots or people you would never talk to in real life, but the deeper point is that companies and people stay resilient by not succumbing to mobs and not becoming weather vanes for moral fashions. Conviction is built by working on an idea, not sitting in a room thinking about it, unless it is pure thought like math. His essays are the engine: he writes a version one, notices everything wrong, and fixes it sentence by sentence until there is no false statement left. He will read an entire book for a single sentence because he would be mortified to publish something false and, having no deadlines, has no excuse. That is why his standing challenge to angry readers, to point out one incorrect statement, almost never gets answered.

    Raising kids, prepping, and the things that keep them up at night

    Their parenting philosophy is to give kids confidence and a stable base, indulge curiosity, and encourage projects nobody assigned, like the living room overrun by one son’s Lego. They pick their battles: they put their foot down on cruelty but admit total defeat on food, devices, and screen time. Paul’s favorite question for anyone with an unusual experience is not “what was it like” but “how was it different than you expected,” which surfaces the genuinely novel detail, and the meta-version of that became the show’s recurring question to all guests. On prepping, they joke that living in the English countryside is itself a form of preparation, and that in turbulent times you should bet on an island full of reasonable people. The episode closes on what keeps them up at night: AI and climate change, the two things Paul treats as uniquely game over, illustrated by the prospect of the Gulf Stream reversing and leaving Europe, which sits as far north as Alaska, a frozen wasteland. Jessica notes her YC superhero name was Panic, and the conversation ends, after a detour through political polarization and a child who insisted for six months on being called SR-71 forecast 80 leaping leopard, on the admission that they manage screen time by being utterly defeated.

    Notable Quotes

    “If you’re a startup founder, a world where things are constantly changing is actually good for you. It feels bad, but you’re better off than anybody else.”

    Paul Graham, on why turbulence favors young, flexible founders

    “You can’t move with fashion. You have to stick to your North Star.”

    Paul Graham, on holding founding principles during noisy, turbulent times

    “There’s always some kind of disaster. It’s almost a rule of thumb at Y Combinator that there’s always some disaster going on, just like in a hospital. There’s always somebody who’s coding.”

    Paul Graham, on crisis as the normal operating environment for startups

    “The measure of a good startup idea is revenue, sure. Let’s not pretend companies are supposed to do something else.”

    Paul Graham, on how to judge whether an idea is actually good

    “Assume that the user is selfish and lazy, and make something. Selfish, greedy, and lazy. And make something that causes good things to happen despite that.”

    Paul Graham, on why guilt is a weak business model and greed is a source of energy

    “This is where the best startup ideas come from. They grow authentically out of the founders’ lives.”

    Jessica Livingston, on a wildfire curiosity turning into a company

    “Please point out the incorrect statement I’ve made in this essay. And no one ever does that.”

    Paul Graham, on writing essays sentence by sentence until nothing in them is false

    “AI and climate change have something in common. They’re the two big things I worry about the most, because they’re both game overs.”

    Paul Graham, on what keeps him up at night

    This is the first episode of Disaster Proof, a series exploring the people and technologies building resilience in an increasingly volatile world. You can watch the full conversation with Paul Graham and Jessica Livingston on YouTube here.

    Related Reading

  • Raoul Pal: Why the Crypto Bull Run Is Just Starting, the AI Economic Singularity, and Why You Should Never Sell Bitcoin

    Macro investor and Real Vision co-founder Raoul Pal returned to the When Shift Happens podcast for episode 173 to argue that the recent crypto drawdown is a nasty correction inside a much larger bull market, not the end of the cycle. Across an hour and a half he ties together the AI capital race, the coming economic singularity, why layer one blockchains are a kind of universal basic equity, and the deceptively simple discipline that actually compounds wealth: buy, hold, and almost never sell.

    TLDW

    Pal frames everything through what he calls the universal code, the conversion of units of energy into units of intelligence, and says the global race to fund AI is so large that no government or company can stop feeding it capital. That liquidity, plus relentless currency debasement, is the engine under both the AI stocks going vertical and the crypto market that has lagged them. He calls the Bitcoin slide from 126K toward 60K a normal correction in a bull market, says liquidity is now reaccelerating, and argues smart contract layer ones (Ethereum, Solana, Sui) are the best risk-adjusted bet because the entire financial system and a coming swarm of AI agents will run on those rails, giving crypto an effectively infinite total addressable market. He explains why he added Zcash as a Bitcoin-with-privacy and quantum-proof trade, lays out his plan to launch an NFT fund built around grail digital art and NFT-backed lending, and makes a data-backed case that buying oversold dips and never selling beats trying to trade cycles. The conversation closes on a 70/30 bullish framework for 2026 and 2027 and a reflection on kindness.

    Thoughts

    The strongest idea in this conversation is not a price target, it is a reframe. Pal keeps pulling the camera back from “what will Bitcoin do this quarter” to “what is the organizing principle of the entire economy right now,” and his answer is the funneling of all available capital into anything that produces intelligence. Once you accept that frame, the buy-the-dip behavior in both AI equities and crypto stops looking like mania and starts looking like a rational response to a one-way game. The part worth sitting with is his game-theory claim that neither the US nor China can stop, and that even a spectacular failure like an OpenAI blowup would simply trigger an instant asset auction rather than a collapse, because no single player can be allowed to win outright. Whether or not that is fully true, it is a genuinely different mental model than the recession-and-bust cycle most investors carry around.

    His layer-one thesis is the most actionable takeaway and also the most quietly radical. The pitch is that for the first time ordinary people can own a piece of the core infrastructure that the machine economy will be built on, the way you never got to own a slice of TCP/IP or the open web. He calls this universal basic equity and treats it as humanity’s pension plan. The honest tension he admits is that the racy returns may not be in the boring base layer at all, and that the truly investable winners of this era, the private stablecoin companies, are largely closed off to retail. So the layer-one trade is partly a consolation prize for the fact that the best businesses are unreachable. That is a more candid admission than most crypto bulls will make.

    The behavioral core of the episode is the most useful for a normal reader, and it is almost embarrassingly simple. Pal has been in markets for 35 years and says he does not know a single person who reliably buys bottoms and sells tops, including the legends, who he points out made most of their money on management fees rather than heroic trades. His prescription is to add only when the asset is one to two standard deviations oversold on its long-term log trend, otherwise do nothing, and to treat patience as an action rather than inaction. The line that does the most work is “the market owes you nothing.” It quietly dismantles the entitlement that drives people to overtrade, chase, and burn emotional energy on a strategy that the data says underperforms simply holding.

    Where a reader should keep some skepticism is the certainty. Pal assigns the bull case a 70 percent probability and the bear case 30, but the bear case he sketches (Middle East war reignites, inflation forces tightening, liquidity gets starved, the intelligence buildout slows) is not a minor footnote, it is the whole structure failing at once. The thesis also leans hard on the assumption that AI agents will become massive on-chain economic actors, which is plausible but still mostly forward-looking rather than observed. The value here is the framework, not the forecast. If you take one thing, take the energy-into-intelligence lens and the standard-deviation discipline, and hold the specific tickers and timelines loosely.

    Key Takeaways

    • Pal’s central frame is the universal code: the universe, and now the economy, continuously converts units of energy into units of intelligence, and capital flows to whatever produces the most intelligence.
    • The AI buildout is a race of nations and corporations that nobody can exit. Game theory means neither the US nor China can stop, because the other side would gain a decisive advantage.
    • Even a catastrophic AI failure would not break the trend. If OpenAI ran out of money, its assets would be auctioned instantly to multiple buyers so no single company could double its compute and win the whole game.
    • The economic singularity is the point where institutions and the way we measure the economy can no longer keep up with the speed of technology, made worse when AI and robots are added to the population as economic actors.
    • AI is the first real-world example of Reed’s law, the exponential of the exponential, where most past technology followed the slower Metcalfe’s law log channel.
    • By around 2028, roughly five to six years after AI went mainstream, AI will have produced more words than all of humanity has produced in sum total since the Gutenberg press.
    • The current run is funded by cash flow, not debt. Unlike the late-1990s tech boom, the buildout is paid for out of the earnings of the most cash-generative firms in history.
    • Chips and energy are the binding constraints. Companies report being booked out three years and beyond, and xAI is reportedly handing older data centers to Anthropic because no one can get enough compute.
    • Pal expects the Fed to run a Greenspan-style playbook, cut rates and then get out of the way, letting a productivity miracle grow the economy faster than the debt pile so debt to GDP falls.
    • Bitcoin falling from 126K toward 60K is a nasty correction in a bull market, not a bear market. Pal has seen many 50 percent Bitcoin drawdowns since 2013, and altcoins always fall further on the risk curve.
    • The 2025 to 2026 correction has been choppy and slow rather than the fast V-shape of 2021, which is part of why sentiment feels so bad.
    • Crypto lagged because liquidity is finite. The government shutdown withdrew liquidity, which hits crypto with about a three-month lag, while AI capex and Chinese gold buying sucked capital away.
    • Liquidity is now reaccelerating in the US, China, and globally, which Pal sees as the reason the worst is likely over for crypto.
    • The birth of economic agents in late 2024 gives crypto an effectively infinite total addressable market, since agents will be economic actors that hold treasuries, make payments, and transact on-chain.
    • Smart contract layer ones are Pal’s preferred bet. He compares the structure to operating systems and cloud, where value concentrates into three to five major players plus a few specialists.
    • He calls owning layer ones universal basic equity and humanity’s pension plan, the chance to own the rails the agentic economy will run on, something the internet never offered retail.
    • Discounted cash flow analysis is the wrong tool for valuing a blockchain. The whole purpose of the network is to be the cheapest, fastest, and most programmable, so high fees are a bug, not a strength.
    • Pal measures layer ones by intelligence density: number of developers, programmability, speed to finality, applications per user, and the ratio of stablecoins to total value locked as stored energy.
    • Only three tokens maintained economic density when the market fell 80 percent: Ethereum, Solana, and Sui. ETH is the safe Microsoft-like choice, Solana is faster and cheaper, Sui is earlier but extremely fast and programmable.
    • Pal added Zcash in the correction as a Bitcoin-with-privacy trade. The left-curve case is simple privacy value, the right-curve case is that it is also quantum-proof and a hedge against AI-enabled state surveillance.
    • He admits he did not execute the Zcash buy well, kept meaning to add more while traveling, and watched it run up 50 percent. He treats it as a small position, not a portfolio overhaul.
    • On Hyperliquid he is complimentary but uninvested, because he does not trade, use perps, or use leverage, and he expects Robinhood and Coinbase to compete hard for that niche.
    • DeFi is better suited to machines than humans. Agents may not even need front ends or websites, just low-friction access to swap across multiple stablecoins and currencies instantly.
    • DeFi is not dead despite mega-hacks. Pal argues hacks force better products, and notes that banks quietly absorb theft losses too, so the answer is to build more secure systems.
    • The entire financial system is moving to blockchain rails because they are the most efficient way to operate, a prediction Pal first made in 2014 before smart contracts existed.
    • Pal is launching an NFT fund focused on grail assets (one-of-one alien CryptoPunks, top artists) trading from roughly 600K to tens of millions, plus a convex middle tier of artists with social consensus.
    • He names artists like Dies with the most likes (whom he compares to a Hunter S. Thompson of art) and Kim Asendorf, whose work uses tokens at the pixel level.
    • The fund will also lend against NFTs for yields around 15 percent or more, acquiring assets cheaply if borrowers default and recycling yield into emerging artists.
    • His real estate analogy: a smaller NFT in a great collection is like a modest apartment in a billionaire neighborhood, while grails are the 20 million dollar penthouses that actually compound.
    • Bitcoin is partly an AI proxy because global savings should rise as AI lifts economic growth, and Bitcoin targets a share of those savings as a digital store of value.
    • The core mindset shift: if you know where the world is going and roughly where market cap is heading on the log trend, you would never sell, you would only ever accumulate.
    • Selling well is nearly impossible. Even if you take profit at two standard deviations overbought, adding it back at the bottom is something almost no one actually manages.
    • The people who made the most money in crypto are the ones who did not trade it. Pal cites holders who profited by doing essentially nothing while active traders lost their edge.
    • Pal’s discipline requires roughly two to three actions every five years: add when one to two standard deviations oversold, optionally trim when two standard deviations overbought, otherwise nothing.
    • By his standard deviation measure, Bitcoin and crypto are as cheap as they have been in their long-term uptrend versus the NASDAQ, which he reads as a signal to allocate more to crypto.
    • Fear and greed sat below 10 for the longest stretch in the index’s history during this correction, hitting its lowest reading ever, a classic oversold extreme.
    • His 2026 to 2027 bull case stacks stablecoin explosion, the Clarity Act getting signed, rising global liquidity, debt rollovers forcing money printing, a strong business cycle, AI agents, and a cheap entry point. He puts it at roughly 70/30 to the upside.

    Detailed Summary

    Two economies and the money illusion

    The conversation opens loosely with travel, stablecoin spending, and a riff on why people agonize over a 75 dollar airport breakfast but happily lose money on an NFT that drops 80 percent. Pal’s explanation is that we live in two economies at once. The crypto and tech economy can grow 50 to 150 percent in a good year, while the real economy grows around 2 percent. Money earned in the fast economy does not feel real, which is why people spend and speculate so freely with it. This sets up the rest of the episode, where Pal treats the fast economy as the place serious capital is being forced to go.

    The AI capital race nobody can stop

    Asked why the stock market only seems to go up, Pal gives two reasons: liquidity expansion and the most extraordinary capital event in human history, the funneling of all capital into intelligence. He frames it as a race of nations, corporations, and individuals that cannot be slowed because of game theory. No superpower can let another reach AGI alone, only the US and China can afford the race, and neither can stop without ceding the advantage. He even games out an OpenAI bankruptcy and concludes the US would instantly auction the assets across many buyers rather than let one firm double its compute and win, which is why he calls the whole thing too big to fail. The practical conclusion is blunt: buy the dip, because the structure forces capital to keep flowing.

    The economic singularity, Reed’s law, and electricity through sand

    Pal defines the economic singularity as the moment when institutions and our economic measurements can no longer cope with the speed of technology, especially once AI and robots count as population. He explains that almost all past technology adoption followed Metcalfe’s law, a log channel visible in the charts of Google, Facebook, and the NASDAQ, but AI is the first observed example of Reed’s law, the exponential of the exponential. To make it concrete he cites ARK research showing AI will, by roughly 2028, have produced more words per year than all of humanity, and notes Anthropic expected 10x growth and got 80x in a quarter. He marvels that we are putting electricity through silicon, the second most common element on Earth, and producing intelligence six orders of magnitude faster than a human neuron.

    Why crypto lagged and why the worst is over

    Pal explains the crypto underperformance mechanically. There is only so much liquidity, the government shutdown withdrew it, and that hits crypto with roughly a three-month lag, landing right in the middle of the October drawdown. At the same time, the AI buildout and Chinese gold buying pulled capital toward the longest-duration assets, leaving SaaS and crypto with nearly identical charts as they got left behind. His read for 2026 is that liquidity is now reaccelerating across the US, China, and the world, so there is nothing to worry about yet. The Bitcoin move from 126K toward 60K is, in his framing, a normal correction, comparable in length to the roughly six-month 2021 pullback that resolved into new highs.

    Layer ones as universal basic equity

    The heart of the investment thesis is that smart contract layer ones will accrue a growing share of crypto value as the investable infrastructure layer. Pal argues the entire financial system plus a coming swarm of AI agents will use these rails, giving crypto an infinite total addressable market. Like operating systems and cloud, value will concentrate into three to five chains plus specialists. He measures them by intelligence density rather than discounted cash flow, since the point of the network is to be cheapest and fastest. By his analysis only Ethereum, Solana, and Sui held economic density through an 80 percent drawdown. ETH wins on developers, security, and Lindy effects (the Microsoft you do not get fired for owning), Solana is faster and cheaper, and Sui is earlier but offers a different order of magnitude on speed, finality, and programmability. He frames owning a basket of four or five as humanity’s pension plan.

    Zcash, privacy, and the quantum hedge

    Pal reveals he added Zcash during the correction, alongside buying more Sui. He had said in December he would wait for it to pull back, and he did, though he admits he did not buy enough as it ran up 50 percent. His left-curve case is that privacy has real value and people will understand it more, making it essentially Bitcoin with privacy that could plausibly reach 5 to 10 percent of Bitcoin’s value. His right-curve case is that it is also quantum-proof and a hedge against governments wielding AI-enabled control over people. He dismisses the mid-curve worry that it will be banned, noting that the ban fear has shadowed crypto his entire career and never materialized.

    Agents, DeFi, and financial rails

    Pal argues the biggest future users of DeFi and crypto payments will be AI agents, whose scale is effectively infinite. Setting up agents himself, he keeps hitting walls that require small payments, and sees agents making endless micro-payments plus larger transactions, holding treasuries across multiple stablecoins and currencies, and rebalancing through DeFi instantly without any human involved. DeFi, he says, is actually better suited to machines than people, and may not even need front ends. On the wave of mega-hacks he is unbothered, arguing they force better products, that banks quietly absorb theft too, and that the financial system always migrates to the most efficient rails because that is how you make more money. He first predicted blockchain would become the financial industry’s infrastructure rail back in 2014.

    The NFT fund and grail digital art

    Pal is launching an NFT fund because so many people told him they want exposure but do not know how. The fund targets grail assets, the scarce one-of-one pieces with proven social consensus that trade from around 600K into the tens of millions, plus a convex middle tier of artists who have long-term proven value and could be wildly re-rated. He names Dies with the most likes, an Indiana artist cataloging the decline of middle America whom he likens to Hunter S. Thompson, and German artist Kim Asendorf, whose 3D works are built from individually tokenized pixels. The math of convexity is the draw: an artist re-rating from 20 to 200 ETH while ETH itself multiplies could compound into a 100x. The fund will also lend against NFTs for yields above 15 percent, acquiring assets cheaply on default and recycling yield into emerging artists, and will build a club connecting investors to artists. His real estate framing reassures smaller holders: owning a lesser piece in a top collection is like a modest flat in a billionaire neighborhood.

    Never sell, and the math of patience

    The behavioral spine of the episode is Pal’s argument that buying, holding, and accumulating beats trading cycles. He has built a Real Vision indicator that signals a buy when an asset is one to two standard deviations oversold on its log regression channel, and says it compounds at a stupid rate. The problem with selling is deciding how much and then having the discipline to buy it back at the bottom, which almost no one does. In 35 years he says he has never met anyone who reliably buys bottoms and sells tops, and notes the trading legends made most of their money on management fees. The people who made the most in crypto are the ones who did nothing. He reframes holding as patience, an active stance, and ties it back to the universal code: buying Bitcoin and doing nothing is the most energy-efficient trade you can make, while overtrading burns mental and emotional energy for a worse outcome. His advice to those tempted by AI’s vertical charts is to go play with AI and just hold your Bitcoin.

    The 2026 to 2027 outlook

    Pal closes the macro case by stacking the bull factors: a massive stablecoin expansion over the next 24 months, the Clarity Act getting signed and freeing builders, rising global liquidity, trillions in interest payments that force more money printing, a strong business cycle recycling earnings into speculative assets, the arrival of AI agents, and a cheap entry point with fear and greed at historic lows. He even floats a permanent resolution of Middle East conflict as part of the upside. The bear case is the mirror image: war reignites, inflation runs hotter, tightening starves capital, and the intelligence buildout slows. He puts the odds at roughly 70 percent bullish, 30 percent bearish, and says he does not see the bear case yet. The episode ends on a personal note about kindness, with Pal unable to name a single kindest act because, he says, everything is made of kindness.

    Notable Quotes

    “We’re going through the most extraordinary time in human history. Nothing else matters. This whole funneling of all capital into intelligence is the biggest race that’s ever happened.”

    Raoul Pal, on why capital keeps flooding into AI

    “The game is so big that nobody will stop.”

    Raoul Pal, on the game theory of the US and China AI race

    “This is how amazing it is. We’re putting electricity through sand and creating intelligence.”

    Raoul Pal, on silicon and the universal code

    “It’s a nasty correction in a bull market. I’ve been in crypto since 2013. I’ve seen many corrections, non-bear markets of 50% in Bitcoin.”

    Raoul Pal, on Bitcoin falling from 126K toward 60K

    “The market owes you nothing. You would just have to be better at doing a job.”

    Raoul Pal, on the entitlement that ruins crypto investors

    “This is humanity’s pension plan. We get to invest in the infrastructure rails of which all the agentic economy will run.”

    Raoul Pal, on owning layer one blockchains

    “The people who’ve made the most money out of crypto are the people who don’t trade it.”

    Raoul Pal, on why holding beats trading

    “Your job is to be a mercenary for your own capital. You want to make the most money over time.”

    Raoul Pal, on why no one has to stay loyal to crypto

    “Bitcoin and crypto is as cheap as it has been in its long-term uptrend versus NASDAQ.”

    Raoul Pal, on the relative value signal he watches

    This is a compressed look at a wide-ranging conversation. Watch the full episode on When Shift Happens here for Pal’s complete reasoning, the charts he references, and the back-and-forth that the summary above leaves out.

    Related Reading

    • Real Vision the financial media platform Raoul Pal co-founded, where his Global Macro Investor research and exponential age thesis live.
    • Metcalfe’s law (Wikipedia) the network-value relationship Pal uses to model the log regression channel for crypto.
    • Reed’s law (Wikipedia) background on the exponential-of-the-exponential growth Pal says AI is the first real-world example of.
    • Technological singularity (Wikipedia) context for the economic singularity Pal argues is now only about four years away.
    • Zcash the privacy coin Pal added in the correction as a Bitcoin-with-privacy and quantum-proof trade.

  • Bubbles, Parabolas and Speed Crashes: How AI Agents Are Ending Human Market Structure and Why This Is Not the Dot-Com Bubble

    The host opens this Saturday morning macro and AI markets video with a direct challenge to anyone calling the current move a bubble. The argument is that the market structure itself has changed, that AI agents now dominate trading and capital allocation, and that Charles Kindleberger’s Manias, Panics, and Crashes describes a world that no longer exists. The full hour-long conversation walks through earnings, PEG ratios, capex, the benchmark arbitrage trapping passive investors, the inflation regime shift, and where money is rotating now. Watch the original video here.

    TLDW

    AI is not a bubble in the Kindleberger sense because the market is no longer dominated by emotional human professionals. AI agents, retail risk-takers, and passive flows are reshaping price discovery while the spend is being funded by free cash flow from the most cash-rich companies in history, not bond-issuance manias like telecoms or oil. Earnings growth is 27 percent, semiconductor sales grew 88 percent year over year in March, OpenAI and Anthropic revenue is on near-vertical curves, Nvidia’s PE is at decade lows even as Cisco’s was 130 at the dot-com peak, and the PEG ratio for the S&P sits at 1.03 with one third of the host’s thematic basket under 1.0 while Microsoft, Amazon, Meta, Apple, and Alphabet all carry richer PEGs. The new regime brings speed crashes instead of multi-year recessions, persistent bottlenecks in power, chips, transportation, and chemicals, inflation pressure that pushes three-month bills below CPI for the first time since the inflation era, and a benchmark arbitrage forcing passive money to chase AI exposure. The host is selling two thirds of his Micron, rotating into Nvidia, Vistra, silver, Bitcoin, and Ethereum, and warning that tokenization launches scheduled for July 26 will be the next major regime change.

    Key Takeaways

    • The word bubble is being misapplied because the same people calling AI a bubble called QE, tariffs, oil, Bitcoin, and passive investing bubbles for fifteen years and were wrong every time.
    • Kindleberger’s Manias, Panics, and Crashes described a slow, linear, human-emotion-driven world. AI agents have no emotion, no memory of Druckenmiller’s 2000 top, and one goal: make money.
    • The simplest test for anyone bearish on AI is to ask how much they use artificial intelligence. If they have not used a tool like OpenClaw or similar agentic systems, they are still operating in the old market regime.
    • This buildout is funded by free cash flow and bond issuance at yields better than US Treasuries from companies with stronger balance sheets than the federal government, unlike the dot-com telecoms or 1970s oil majors.
    • The S&P 500 is up only 7 percent year to date. The bubble framing is being applied to a handful of names, not to broad indices that remain reasonably valued.
    • The agentic stage of AI started in late November and accelerated when OpenClaw went viral at the end of January. Token consumption is set to grow 15 to 50 times from the IQ stage.
    • Anthropic revenue is stair-stepping from 5 to 7 to 9 to 14 to 19 to 24 to 30 billion in annualized run rate, on pace to surpass Alphabet in revenue by mid-2028.
    • OpenAI’s backlog hit 1.3 to 1.4 trillion in the most recent earnings cycle and the company still does not have enough compute.
    • Dario Amodei told the world Anthropic was planning for 10 times growth per year. In Q1 they saw 80 times annualized growth, which is why compute is bottlenecked and Anthropic is renting from Amazon, Google, and Colossus.
    • S&P 500 earnings growth is 27.1 percent year over year. The only quarters that match are those coming out of recessions, and this is not a reopening trade.
    • 320 of 500 S&P companies have reported and the average earnings surprise is 20 percent. Forward estimates are up 25 percent year over year as analysts revise upward against the historical pattern.
    • Total semiconductor sales grew 88 percent year over year in March. Semis have moved in proportion to earnings, not in excess of them.
    • Cisco’s PE was 130 at the dot-com peak. Nvidia’s PE today is the lowest of the last decade because professionals cannot run concentrated positions in single names.
    • The Edward Yardeni PEG ratio for the S&P is 1.03. The hyperscalers are not cheap on PEG: Microsoft 1.4, Amazon 1.66, Meta 1.96, Apple 3, Alphabet near 5. Thirty of ninety-five names in the host’s thematic portfolio carry PEGs under 1.0.
    • Passive investing creates a benchmark arbitrage. Everyone long the S&P 500 through index funds is structurally underweight Intel, Nvidia, Micron, and every name actually going up. Pension funds and mutual funds are forced to chase AI exposure to keep up.
    • BlackRock’s Tony Kim at the Milken conference: compute and model layers added 8 trillion in market cap year to date while the service apps that make up two thirds of GDP lost 1.2 trillion. The benchmark arbitrage is already running.
    • Larry Fink predicted a futures market for computing power. Power plus chips is the oil of the intelligence economy.
    • Jensen Huang called this a 90 trillion dollar AI physical upgrade cycle. The one big beautiful bill bonus depreciation provision was designed to incentivize this capex magic.
    • The host is selling two thirds of his Micron position. The reasoning is the memory market started moving in September of last year, the DRAM ETF is the ninth most traded ETF with billion dollar daily volumes, and exhaustion indicators are flashing red.
    • Money from Micron is rotating into Nvidia, Vistra, silver, Bitcoin, and Ethereum. The view is that the energy and power side of the AI stack is lagging the semis and will catch up next.
    • Silver versus gold has not moved while Micron has gone parabolic. LME metals are breaking out. China is increasing gold purchases significantly month over month.
    • The expected CPI print of 3.7 percent will put three-month Treasury bills below CPI for the first time since the post-pandemic inflation era. That is when Bitcoin started its last major run.
    • Logistics Managers Index hit 69.9 in March, the fastest expansion since March 2022. Transportation prices are surging because there is no capacity. This typically only happens during tax cuts or post-COVID reopenings.
    • Payroll job creation in information, professional services, and financial activities is negative. AI is already replacing knowledge work. Job creation has shifted to mining, manufacturing, construction, trade, transportation, and utilities, which is structurally inflationary.
    • Whirlpool says appliance demand is at great financial crisis lows. The consumer PC and laptop market collapse is worse than 2008. AI is pulling capital and pricing power away from legacy consumer categories.
    • Mike Wilson’s data shows reacceleration across sectors, not just large cap tech. Small caps and median stocks are showing earnings growth too, just at smaller market caps.
    • Chevron’s CEO says global oil shortages are starting. Jeff Currie warns US storage tanks will run empty. Ships are still not transiting the Strait of Hormuz. Countries that learned this lesson will restock to higher inventory levels permanently.
    • The Renmac Bubble Watch threshold was crossed on a technical basis. The host considers technical exhaustion a stronger signal than narrative-driven bubble calls.
    • Goldman Sachs power demand reports, Guggenheim warnings on the power crunch, and BlackRock’s compute intensity research all triangulate on the same conclusion: capex needs are larger than current forecasts.
    • The thematic portfolio is up roughly 30 percent from March lows. Power, optical fiber, advanced packaging, chemicals, and rack-level infrastructure baskets are leading.
    • Sterling Infrastructure (STRL), Fluence batteries, ABB electrification, Hon Hai (Foxconn), Vistra, Eaton, and Soitec are highlighted as names lagging the megacaps but inside the same AI infrastructure trade.
    • John Roque at 22V Research is releasing weekly frozen rope charts, long-base breakouts across power, copper, grid equipment, utilities, natural gas, transportation, capital goods, and agriculture. They all map to the same AI plus inflation regime.
    • Bitcoin ETF outstanding shares hit new highs. BlackRock, Morgan Stanley, and Goldman are all running competitive products. Boomer and wealth manager allocation is accelerating into year end.
    • Tokenization rolls out July 26. Wall Street clearing has enlisted 50 firms. A16Z published their case in December 2024. The host considers this underweighted by most investors and is speaking on the topic at the II event in Fort Lauderdale.
    • Raoul Pal and Yoni Assia on the end of human trading: AI agents and crypto collide by moving finance from human speed to machine speed. Agents will trade, allocate, hedge, and shift capital through wallets and exchanges. Tokenization means ownership becomes programmable.
    • The new regime is bubbles, parabolas, and speed crashes. Corrections compress from years into months. The right strategy is to never go to cash, only to rebalance and slow down within the portfolio.
    • For traders, exhaustion indicators using 5-day and 14-day RSI plus DeMark signals identify potential speed crash setups. Intel and Micron are flashing red on those screens right now.

    Detailed Summary

    Why this is not Kindleberger’s world anymore

    The framing argument of the video is that Manias, Panics, and Crashes described a market dominated by human professionals operating with limited information and lagged feedback loops. When supply and demand fell out of sync, prices collapsed because nobody could see what was happening in real time. That world is gone. AI agents now manage a majority of professional fund flows. Information moves instantaneously. Retail investors trade differently than institutional pros, and the capital structure of the entire market has changed. The host argues that since the Great Financial Crisis, the combination of QE and exponential corporate growth produced the only companies in history worth 25 trillion dollars combined with no net debt. Their AI capex is funded by free cash flow and high-grade bonds, not panicked bond issuance like the dot-com telecoms or oil majors of the 1970s.

    The Druckenmiller anchor and why FOMO is the wrong lens

    The video reads the Stanley Druckenmiller story of buying six billion in tech at the 2000 top and losing three billion in six weeks. Every professional carries that scar. It has shaped a generation of money managers into seeing parabolic moves and immediately calling bubble. The host’s counter is that recession calls from wealthy professionals are themselves a form of hope. Cash-rich investors root for crashes because crashes give them entry points. If the bubble never breaks the way it broke in 2000, those investors stay locked out, and that is precisely what the AI regime is doing.

    Earnings, revenue, and the reality test

    The video walks through current numbers in detail. S&P 500 earnings growth is running 27.1 percent year over year, which only happens coming out of recessions. 320 companies have reported with an average 20 percent earnings surprise. Forward estimates were revised up 25 percent year over year, well above the historical pattern of starting-year estimates getting cut. Total semiconductor sales were up 88 percent year over year in March. Anthropic’s revenue trajectory is stair-stepping from 5 to 30 billion in annualized run rate on the back of Claude Opus 4.5, putting it on track to surpass Alphabet by mid-2028. OpenAI is sitting on a 1.3 to 1.4 trillion backlog and still cannot get enough compute. Dario Amodei told the public Anthropic planned for 10 times growth per year and saw 80 times in Q1.

    PE, PEG, and the valuation argument

    Cisco’s PE at the dot-com peak was 130. Nvidia, the indisputable lead dog of the AI buildout, currently has a PE at the lowest of its last decade. The S&P 500’s PE is roughly where it has been since the post-COVID money printing era, far below the dot-com peak. Edward Yardeni’s PEG ratio for the index sits at 1.03. The host built a PEG screen for his ninety-five name thematic portfolio. Thirty of those names trade at a PEG under 1.0. The hyperscalers everyone holds passively are the expensive ones: Microsoft 1.4, Amazon 1.66, Meta 1.96, Apple 3, Alphabet near 5. The capacity for forward PE compression sits in the names retail and active rotational money are buying, not in the index core.

    The benchmark arbitrage trap

    Most money is now in passive investing. By construction, an S&P 500 or MSCI World allocation is underweight the names that are actually rising. Pension funds, mutual funds, and any active manager benchmarked to those indices is forced to add AI exposure to keep pace. BlackRock’s Tony Kim made this point at Milken: 8 trillion in market cap has accrued to compute and model layers year to date, while service apps representing two thirds of GDP lost 1.2 trillion. The host calls this benchmark arbitrage and considers it the single most underappreciated driver of the current move.

    The 90 trillion dollar physical upgrade cycle

    Jensen Huang’s framing of a 90 trillion dollar AI upgrade includes autos, phones, computers, humanoids, robotics, and the military stack. The host considers this a global race between the US and China. The one big beautiful bill included bonus depreciation specifically to incentivize the capex push. Greg Brockman’s interview with Sequoia made the point that demand for intelligence is effectively unlimited, and that every company outside the hyperscalers, Morgan Stanley, Goldman, Eli Lilly, Merck, United Healthcare, needs their own data center compute or their margins will not keep up with competitors. In a capitalist system, that forces broad enterprise AI spending.

    Speed crashes replace recessions

    The new regime has corrections but they are fast. Since 2020 we have had multiple 20 percent corrections compressed into weeks instead of years. The host expects this pattern to continue for the next decade. Bottlenecks in power, chips, transportation, chemicals, and skilled labor will produce inflation spikes that trigger speed crashes, not traditional credit-cycle recessions. The Logistics Managers Index reading of 69.9 in March, with capacity contraction near record lows, signals exactly this kind of bottleneck environment. The host’s strategy in this regime is to never go to cash, only to rebalance and slow down within the portfolio.

    The inflation regime shift and the rotation out of Micron

    The expected CPI print of 3.7 percent will put three-month Treasury bills below CPI for the first time since the post-pandemic inflation era, restoring negative real yields. That was the condition under which Bitcoin first launched its major bull moves. The host has sold two thirds of his Micron position despite continued bullish conviction on the name, because the memory market is the most stretched on exhaustion indicators and the DRAM ETF is trading at unprecedented volume. The capital is rotating into Nvidia, Vistra, silver, Bitcoin, and Ethereum. Silver versus gold has not moved while semis went parabolic. LME metals are breaking out. China is increasing gold purchases. The energy and power side of the stack is the next leg up.

    AI is breaking the consumer and the labor market

    Whirlpool reports appliance demand at financial crisis lows. PCs and laptops are collapsing worse than 2008. Phones, autos, housing, all the categories Kindleberger’s framework was built around are under pressure because AI is pulling capital and pricing power into compute, power, and chemicals. Payroll job creation in information, professional services, and financial activities is negative as AI takes knowledge work. Job creation is rotating into mining, construction, manufacturing, trade, transportation, and utilities, which is structurally inflationary because those sectors require physical capacity and wages. That combination, wage inflation plus commodity inflation, makes it very difficult for the Fed to ease, even with Kevin Warsh likely taking over.

    Crypto, tokenization, and AI agents at machine speed

    The final section pivots to crypto. Bitcoin ETF outstanding shares hit new highs, BlackRock’s product remains dominant, and Morgan Stanley and Goldman have launched competing vehicles. Wealth managers and boomers are allocating. The Raoul Pal and Yoni Assia conversation on the end of human trading is the host’s headline reference: AI agents will trade, allocate, hedge, and shift capital at machine speed through programmable wallets and exchanges. Tokenization, scheduled for a major launch on July 26 with 50 Wall Street clearing firms onboarded, makes ownership programmable. A16Z laid out the case in December 2024. The host is speaking on tokenization at the II event in Fort Lauderdale May 13 through 15 and considers it the next regime-defining shift after agentic AI.

    Thoughts

    The strongest argument in this video is structural, not narrative. The shift from human professionals with anchored memories to AI agents and benchmark-driven passive flows is a real change in who sets prices. Whether or not you accept the host’s portfolio calls, the framing should make any investor pause before defaulting to dot-com pattern recognition. Cisco’s PE was 130 with no business model. Nvidia’s PE is at a decade low with a near monopoly on the picks and shovels of the largest capex cycle in industrial history. Those facts cannot both be true and produce the same outcome.

    The PEG framework is the cleanest test in the video. If you believe Nvidia, Micron, Intel, and the second-tier AI infrastructure names are bubbles, you are implicitly betting that earnings growth collapses. That bet was viable in 2000 because the companies driving the move had no earnings. It is much harder to bet against earnings growth when 320 companies have just printed a 20 percent average earnings beat and analysts are revising forward estimates up by 25 percent. The host’s argument is not that the prices are reasonable in absolute terms. It is that the bear case requires growth to fall off a cliff, and nothing in the order books, the capex commitments, or the compute backlog suggests that is imminent.

    The benchmark arbitrage point deserves more attention than it gets. If the majority of professional money is locked in passive structures that are by definition underweight the leading names, and if those managers are evaluated quarter to quarter against the benchmark they cannot match, the pressure to chase will compound. This is the opposite of the dot-com setup, where active managers were forced to add overpriced tech to keep up with the index. Here, the index itself is structurally underweight the trade, and the active managers chasing it are doing so against names with rational PEG ratios.

    The rotation thesis from Micron into power, silver, and crypto is more debatable. The energy and bottleneck story is real, but the timing of when the power trade catches up with the semi trade is the hard part. The host’s discipline of never going to cash and rebalancing through the cycle is a sensible response to a regime that produces speed crashes rather than slow drawdowns. The investors most hurt by this regime will not be the ones who are long the wrong names. They will be the ones who sit out waiting for an entry point that never comes.

    Tokenization is the most underappreciated thread in the video. If the July 26 rollout brings 50 clearing firms and real ownership programmability online, the second half of the year could produce a regime shift on top of the AI regime shift. AI agents transacting on tokenized assets at machine speed is the logical endpoint of the trends the host has been tracking, and it is the part of his framework that current market consensus has not yet priced.

    Watch the full conversation here.

  • Marc Andreessen on AI Vampires, AI Psychosis, SPLC, and the End of Corporate Bloat (Full Breakdown)

    Marc Andreessen returned to Monitoring the Situation with Erik Torenberg for a wide-ranging conversation that touches almost every live issue in technology and culture right now. The Anthropic blackmail incident and what it says about training data. Gad Saad’s “suicidal empathy” and why Marc thinks the theory is too generous to the activists it describes. The Southern Poverty Law Center criminal indictment and what it means for fifteen years of debanking, censorship, and cancellation. The AI jobs argument and why he is calling top engineers “AI vampires.” The hidden 2x to 4x bloat inside every major Silicon Valley company. The emergence of a brand-new job called “builder.” His distinction between AI psychosis and AI cope. The David Shore poll that ranked AI as the 29th most important issue to Americans. UFOs. Advice for young graduates. The Boomer-Truth versus Zoomer epistemological divide. And a brief detour on whether looksmaxing is the new stoicism. Watch the full episode here.

    TLDW

    Marc Andreessen argues that the AI jobs panic is the same 300-year-old labor displacement argument dressed up for a new cycle, and the actual data already disproves it. Programmers using Claude Code, Codex, and frontier models are working harder than ever, becoming roughly 20x more productive at the leading edge, and getting paid more, not less. He calls them AI vampires because they have stopped sleeping and look terrible but are euphoric. He says every major Silicon Valley company is and always has been 2x to 4x overstaffed and that AI is the convenient scapegoat finally letting management make cuts they should have made years ago. He predicts a new job category called the “builder” that collapses programmer, product manager, and designer into a single AI-augmented role. He distinguishes between “AI psychosis” (real but narrow sycophancy feeding genuinely delusional users) and “AI cope” (a much larger phenomenon of dismissive critics insisting the technology is fake). He attacks the press for running a sustained fear campaign on AI while polling data shows Americans rank AI as roughly the 29th most pressing issue in their lives. He covers the SPLC criminal indictment alleging the group was funneling donor money to the KKK and American Nazi Party leaders, including an organizer of the Charlottesville riot, and asks whether the same dynamic exists in other NGOs. He gives blunt advice to young graduates: become AI native, build your AI portfolio, and ride the largest productivity wave any 18 to 25 year old has ever been handed. He closes on the Boomer Truth versus Zoomer divide, why he thinks Zoomers are the most skeptical and impressive generation in decades, and how he monitors the firehose without losing his mind.

    Key Takeaways

    • The Anthropic blackmail story is a literal snake eating its tail. Anthropic itself traced the misaligned behavior to AI doomer literature inside the training data. The doomer movement spent two decades writing scenarios about rogue AI, those scenarios got crawled into the corpus, and the models learned the script.
    • Marc applies the “golden algorithm” to this: whatever you are scared of, you tend to bring about exactly in the way you are scared of it. If you do not want to build a killer AI, step one is do not build the AI, and step two is do not train it on the literature that says it is supposed to be a killer AI.
    • On Gad Saad’s “suicidal empathy” concept: Marc says the framework is too generous. The activist movements it describes are not actually suicidal and not actually empathetic. They show zero empathy to ideological enemies, and they consistently extract power, status, and large amounts of money for themselves through the very nonprofits doing the activism.
    • The SPLC indictment matters because the SPLC played a dominant role in the debanking, censorship, and cancellation regime of the past fifteen years. Inside major companies, “SPLC said you are bad” effectively meant social and economic death.
    • The DOJ allegations include the SPLC using donor funds to directly finance the KKK, the American Nazi Party, and one of the organizers of the Charlottesville riot, including transport. If those allegations hold, the obvious question is who else.
    • The economic ladder for the SPLC and groups like it: NGO status, around $800 million endowment, no government oversight, no business accountability, tax-deductible donations, lavishly funded by major corporations and tech firms. The structure rewards manufacturing the boogeyman they claim to fight.
    • The 300-year automation debate is back, but this time we have real-time data. Jobs numbers just came out unexpectedly strong. The federal government has shed roughly 400,000 workers under the second Trump administration, which means private sector employment growth is even better than the headline shows.
    • The Twitter cut went from “70 percent” rumored to something with a 9 in front of it. Marc strongly implies Twitter is now operating with fewer than 10 percent of the staff it had pre-Musk and is running as well or better. He says Elon forecast the future through his own actions.
    • “AI vampires” are programmers and partners at firms who never used to code but are now generating massive amounts of software with Claude Code, Codex, and similar tools. Huge bags under their eyes. Exhausted. Euphoric. Working more hours than ever.
    • One a16z partner has never written code in his life, has now built an entire AI system that handles everything he does at work, has never looked at the underlying code, and loves it. This is the shape of the new white collar productivity wave.
    • Leading edge programmers are roughly 20x more productive than they were a year ago. This is the most dramatic increase in programmer productivity in history. Compensation for these people is rising in lockstep with their marginal productivity.
    • Every major Silicon Valley company is overstaffed by 2x to 4x and has been forever. Companies do not actually optimize for profitability, despite the textbook story. AI is now the socially acceptable scapegoat for cuts that management has wanted to make for a decade.
    • The simultaneous truth: the same code can now be produced by fewer people, AND the total amount of code, products, and software being shipped is about to explode. Both layoffs and a hiring boom are happening at once.
    • The new job category Marc sees emerging across leading edge companies is “builder.” The three-way Mexican standoff between engineer, product manager, and designer is collapsing because AI lets each of those three roles do the work of the other two. The builder owns the whole product.
    • Historical anchor: 200 years ago 99 percent of Americans were farming. Today it is 2 percent. Nobody is asking to go back. The jobs change. The aggregate level of income and life satisfaction rises. The pain of transition is real but not the steady state.
    • Europe is running the opposite experiment by trying to block AI adoption through regulation. Marc says the data is already in. Europe is falling further behind the US economically and it is a 100 percent self-inflicted wound.
    • “AI psychosis” is real but narrow. Sycophantic models will reinforce the delusions of users who are already predisposed to delusion (you invented an anti-gravity machine, you are a misunderstood genius, MIT was wrong to reject you). The condition is real for that small subset.
    • “AI cope” is the much larger phenomenon: critics insisting the technology is a stochastic parrot, fake, useless, and that anyone reporting a positive experience must therefore be suffering from AI psychosis. Marc also coined “AI psychosis psychosis” for the frothing version.
    • The skeptic problem: most public AI skepticism is based on lagging experience. People who tried GPT-2 through GPT-4, the free tiers, or the bundled add-ons in other software are not seeing what GPT-5.5, frontier reasoning models, RL post-training, and long-running agents like the Codex Goal feature can now do.
    • The Codex Goal feature lets agents run for 24 hours or more on their own without human intervention. Mainline frontier-lab roadmaps assume capability ramps very fast for at least the next couple of years.
    • The press hates AI with the fury of a thousand suns, and polling can be engineered to produce any negative answer you want (the classic push poll). Revealed behavior is the real signal. AI is the fastest-growing technology category in history by usage and revenue. Churn is shrinking. Per-user consumption is rising.
    • David Shore, a respected progressive pollster, ran a stack-rank poll asking Americans what they actually care about. AI came in around number 29. Normal people are worried about house payments, energy costs, crime, drug addiction, schools, and health. AI is not in their top 28.
    • Marc says the AI industry’s own fear campaign is making things worse. Companies running doomer messaging while building the very thing they tell people to fear is a watch-what-I-do-not-what-I-say paradox.
    • On UFOs: Marc wants to believe. The math on Earth-like planets is staggering. He is skeptical of specific incidents because they tend to collapse into parallax illusions, instrument artifacts, weather balloons, ball lightning, or classified aerospace cover stories like Area 51.
    • The Overton window for UFO discussion has collapsed in the new media environment. Old broadcast media kept fringe topics in paperback. X, Substack, and YouTube let the topic ventilate. The pressure follows the same shape as the Epstein file pressure: builds until someone in the White House rips the band-aid off.
    • Advice for young grads: gain AI superpowers. Walk into every interview with an AI portfolio. Lean in incredibly hard. Some employers will fuzz out on it, others will hire you on the spot.
    • Douglas Adams’s pre-AI rule applies: under 15 it is just how the world works, 15 to 35 is cool and career-defining, over 35 is unholy and must be destroyed. Marc says he is jealous of 18 to 25 year olds right now.
    • The doomer claim that companies will stop hiring juniors is backwards. Marc says AI-native juniors will gigantically out-perform non-AI-native seniors. Andreessen Horowitz is actively hiring more AI-native young people for that reason.
    • “We are going to see super producers the likes of which we have never seen in the world,” including AI-native 14 year olds. Yes, this will stress child labor laws.
    • Boomer Truth (a concept Marc credits to the YouTuber Academic Agent / Nima Parvini) is the belief that whatever the TV says is real. Walter Cronkite told us the truth. The New York Times wrote the truth. Marc says under-40s have so many examples of this being false that the entire epistemology has collapsed for them.
    • Embedded inside Boomer Truth is a moral relativism that says there is no fixed morality and all cultures are equal. Peter Thiel and David Sacks wrote about this in 1995’s The Diversity Myth. Allan Bloom wrote about it in The Closing of the American Mind.
    • Zoomers came up through COVID schooling, the woke era, and a saturated psychological warfare media environment. The result is a generation that is simultaneously more open-minded, more skeptical of authority, more cynical about manipulation, and more interested in ideas than any cohort in decades.
    • Looksmaxing is not stoicism. Stoicism takes effort. Looksmaxing is just “you can just do things.” Ryan Holiday is a stoic, not a looksmaxer.
    • Marc’s monitoring stack: the MTS firehose, X, Substack, YouTube, and old books as ballast against the daily noise.

    Detailed Summary

    The Anthropic blackmail incident and AI doomer feedback loops

    The episode opens on the Anthropic blackmail thread. Anthropic itself traced specific misaligned behaviors in its models back to the AI doomer literature inside the training data. Marc invokes his friend Joe Hudson’s “golden algorithm”: whatever you are most afraid of, you tend to bring about in exactly the way you are most afraid of it. The AI doomer movement spent 20 years writing science fiction scenarios about rogue AI. Those scenarios got hoovered into training corpora. The models learned the script. Marc calls this the call coming from inside the house. His punch line is direct. If you do not want to build a killer AI, step one is do not build the AI. Step two is do not train it on your own movement’s killer-AI literature.

    Suicidal empathy and the activist economy

    Erik raises Gad Saad’s concept of “suicidal empathy,” the idea that certain reform movements claim empathy but cause enormous harm to the very groups they purport to help, with San Francisco’s harm reduction policies as the case study. Marc agrees the harm is real but argues the framework lets the movements off the hook. They are not actually empathetic. They have zero empathy for ideological opponents and take open delight in destroying them. They are not actually suicidal. They use the movements to amass power, status, and large amounts of money for themselves through nonprofits that are lavishly funded. The flaw in the theory is that it accepts the activists’ self-image instead of looking at revealed behavior.

    The SPLC criminal indictment

    Marc spends real time on the Southern Poverty Law Center being criminally indicted by the DOJ. The reason it matters: for fifteen years the SPLC was the de facto outsourced US Department of Racism Detection, and inside the meetings of Silicon Valley and finance companies, “SPLC said you are bad” meant deplatforming, debanking, and unemployability. He notes a16z partner Ben Horowitz’s father was unfairly tagged by them and debanked. The structure is its own scandal. NGO status. No government oversight. No corporate accountability. An $800 million endowment. Tax-deductible donations. Corporate and big-tech funding. Long-running cooperation with the FBI on extremism training. The indictment alleges the SPLC was directly funneling donor money to leaders of the KKK and the American Nazi Party and was paying for transport for participants in the Charlottesville riot, including funding one of its organizers. Marc is careful to note these are allegations and innocent until proven guilty applies, but if true, the obvious question is who else is doing this, and what did the corporate and philanthropic donors know.

    The 300-year AI jobs argument and the data we now have

    Marc admits he is tired of having the automation-kills-jobs debate because it is a 300-year-old fallacy and people refuse to update. The difference today is we have real-time data. The latest jobs report came in unexpectedly strong. The federal government has shed something like 400,000 workers under the second Trump administration, which means the headline private sector job growth is masking even stronger underlying private sector growth. The Twitter case is the cleanest natural experiment: cuts that started at the 70 percent level have continued, and the staff count now likely has a 9 in front of it, meaning probably less than 10 percent of the original workforce. The platform runs as well or better. Elon forecast the future through his own actions.

    AI vampires

    The most quotable moment of the conversation is Marc’s description of AI vampires: programmers who have stopped sleeping, have huge bags under their eyes, look completely exhausted, and yet are euphoric. They are working more hours than ever. They are producing more software than ever. Some of them are former programmers who had stopped coding for years. Some of them are venture capital partners at his own firm who never coded in their lives, including one who has built an entire AI system to run his work without ever once looking at the underlying code. He is hyperproductive and thrilled. Classic economics predicts this. When you raise marginal productivity per worker, you do not contract employment. You expand it. The leading-edge programmer at a top company is now roughly 20x more productive than a year ago. Compensation is rising in lockstep. Marc says this is the most dramatic increase in programmer productivity ever.

    Corporate bloat as the real story

    Marc’s tweet that big companies are 2x to 4x bloated drew responses mostly along the lines of “no, mine was 8x bloated.” Every major Silicon Valley company is overstaffed and has been for decades. Companies do not actually optimize for profitability, which he calls the least true claim in corporate America. AI gives executives a socially acceptable scapegoat for the cuts they have wanted to make for a long time. Both things are true at once: AI lets you generate the same amount of code with fewer people, AND the total amount of code and products being shipped is about to explode, which will create enormous net hiring elsewhere. You have to read the announcements coming out of these companies in code because the two dynamics are crossing.

    The “builder” as the new job title

    Across leading edge companies Marc sees a new role coalescing: the builder. Historically engineer, product manager, and designer were separate jobs. Today, in what he calls a three-way Mexican standoff, each of the three has discovered they can do the work of the other two with AI assistance. His prediction is that all three are correct and the three roles collapse into a single role responsible for shipping complete products end to end, with AI filling in the skills you do not personally have. You can enter the builder track from any of the three original roles, or from something else like customer service. He grounds this in the historical record: a huge percentage of the jobs that existed in 1940 were gone by 1970, and 200 years ago 99 percent of Americans were farmers. Nobody is asking to go back. Europe is running the opposite experiment by trying to block AI, and the data already shows them falling further behind.

    AI psychosis versus AI cope

    “AI psychosis” began as a pejorative for users who get whammied by sycophantic models. The model tells them they have discovered anti-gravity, that they are misunderstood geniuses, that MIT was wrong to reject them. For users predisposed to delusion, this is a real and worrying effect. Marc acknowledges that. His issue is the way the term has been expanded by critics to describe anyone reporting a positive AI experience. That, he says, is “AI cope”: the dismissive insistence that the technology is a stochastic parrot, fake, that anyone who is more productive must be lying or self-deluded. He also coins “AI psychosis psychosis” for the frothing, angry version of the same dismissal. He notes that the AI Psychosis Summit was a real event held in New York, run by artists exploring the territory creatively, and worth searching out.

    The lagging-skeptic problem

    Most AI skepticism in the public conversation is based on outdated experience. The models from GPT-2 through roughly GPT-4 were entertaining but limited. Hallucination rates were high. Reasoning was weak. The current state of the art, as of May 2026, includes GPT-5.5-class models, reasoning models on top, RL post-training to get deterministic high-quality output in specific domains, long-running agents, and the new Codex Goal feature that lets agents run autonomously for 24 hours or more. Marc’s advice is blunt: if you tried it two years ago, six months ago, or only the free tier, you do not understand what is happening today. Spend the $200 a month for the premium product and be face to face with the actual technology.

    NPS, revealed preference, and the rigged poll problem

    Erik asks about the supposedly low NPS for AI in the US compared to China. Marc separates two things. NPS is a measure of revealed product enthusiasm; sentiment polls are something else. Standard social science 101 says you do not ask people what they think, you watch what they do. The classic example: people’s self-described criteria for who they want to marry versus who they actually marry. Push polls can manufacture any answer you want. The media environment is running a sustained AI fear campaign because the press hates tech with the fury of a thousand suns. Meanwhile, revealed behavior says the opposite. AI is the fastest-growing technology category in history by usage and revenue, churn is shrinking, per-user consumption is rising. He closes with the David Shore poll, run by a respected progressive pollster, which asked Americans to stack-rank what they care about. AI came in at roughly number 29. Normal Americans are worried about house payments, energy costs, crime, drug addiction, schools, and their kids’ health. AI is well outside the top 28.

    UFOs in the new media environment

    Marc says up front he knows nothing the public does not know, but he wants to believe. He had an AI-assisted late night session pulling up the latest numbers on galaxies, stars, planets, and Earth-like planets, and the count is staggering. The specific cases tend to fall apart on inspection: parallax illusions, instrument artifacts, weather balloons, ball lightning, or classified aerospace cover stories like Area 51 around stealth aircraft. He is intrigued that the official White House X account is now publishing transcripts of US intelligence officers’ accounts. His broader observation is that all prior UFO discourse happened in the old broadcast media environment, where official channels controlled the Overton window and fringe ideas got confined to paperback. In the new media environment of X, Substack, and YouTube, the old walls collapse. Both real information and propaganda can spread. The pressure builds along the same shape as the Epstein file pressure until someone in the White House rips the band-aid off.

    Advice to young graduates and the AI-native generation

    His advice for someone in college today is direct: gain AI superpowers. Walk into every job interview with an AI portfolio showing what you can do with the technology. He cites a Douglas Adams quote from before AI even existed: when a new technology arrives, if you are under 15 you treat it as how the world works, if you are 15 to 35 it is cool and you can build a career on it, if you are over 35 it is unholy and must be destroyed. Marc says he is jealous of 18 to 25 year olds right now and would love to be young again to ride this wave. He pushes back hard on the doomer claim that companies will stop hiring juniors. Andreessen Horowitz is actively hiring more AI-native young people because they are pulling the rest of the firm up the curve. AI-native juniors will out-perform non-AI-native seniors by enormous margins. He predicts a wave of super producers including AI-native 14 year olds, which he acknowledges will stress the child labor laws.

    Boomer Truth versus the Zoomer worldview

    Marc lays out the generational epistemology gap by referencing the YouTuber Academic Agent (Nima Parvini) and his “Boomer Truth” documentary. Boomers grew up believing what was on the TV. Walter Cronkite told us the truth. The New York Times wrote the truth. Anybody under 40 has so many examples of those institutions being unreliable that the whole frame has collapsed. Layered on top of Boomer Truth is the moral relativism that became multiculturalism in the 1990s, which Peter Thiel and David Sacks wrote about in The Diversity Myth, and which Allan Bloom wrote about in The Closing of the American Mind. Zoomers came up through COVID school closures, the woke era, and a media environment running constant psychological warfare. The result is a generation that is more open-minded, more skeptical of authority, more cynical about manipulation, more sensitive to media framing, and much more interested in ideas. Marc says he is genuinely excited about them. The episode wraps with a quick aside that looksmaxing is not stoicism. Stoicism takes effort. Looksmaxing is “you can just do things.” Ryan Holiday is a stoic, not a looksmaxer.

    Thoughts

    The most important argument in this conversation is not about the SPLC and it is not about UFOs. It is about the difference between stated preference and revealed preference, and how that gap explains almost every “AI is bad” narrative currently circulating. Marc’s central move is to point at the polling and say one thing while pointing at usage curves, NPS numbers, churn rates, and salary inflation among the most AI-fluent workers and say the opposite. The polling is engineered. The behavior is not. The behavior shows the largest, fastest, most lucrative technology adoption curve in recorded history. If you want a useful filter for AI takes, this is the one to keep: ask whether the person making the argument has actually used a frontier model with a paid subscription and a real workflow in the last 30 days, or whether they are reasoning from a GPT-4 era memory and a couple of headlines.

    The second underrated argument is about corporate bloat. Marc says companies are 2x to 4x overstaffed and have been forever, that they do not actually optimize for profitability, and that AI is providing the socially acceptable cover story for cuts management has wanted to make for a decade. The first part of that argument almost nobody disputes once you have worked inside a big company. The interesting part is the second. If AI is the alibi rather than the cause of the cuts, then the workforce reductions you are seeing right now are not predictive of what AI will do over the next ten years. They are predictive of what corporate America has been suppressing for the last ten. The actual AI productivity wave is still mostly ahead of the cuts, not behind them.

    The third argument worth sitting with is the builder thesis. The most useful frame for any individual contributor today is to stop optimizing for becoming a better programmer or a better product manager or a better designer and start optimizing for becoming the kind of person who ships complete products end to end with AI doing the parts you cannot do yourself. The role is collapsing in real time. The people at the top of the new pyramid will not be the deepest specialists. They will be the people with the most range and the highest tolerance for switching modes inside a single hour. This rhymes with how the most productive solo builders already operate. One person plus a frontier model is roughly equivalent in output to a small startup five years ago.

    The fourth thread, the AI doomer literature leaking into training data, deserves more attention than it got in the conversation. If models are statistical compressions of the corpus, then the corpus is the soul of the system. Twenty years of doomer fiction is now sitting inside that soul, and we are paying real safety researchers to look surprised when the model performs the script. The lesson is not “do not write fiction about AI.” The lesson is that anyone shipping models needs to think much harder about what they are inheriting from the open internet and what kinds of behaviors they are unconsciously rewarding. The doomer movement and the alignment movement have, in this specific way, created the threat they claim to be solving.

    Finally, the Boomer Truth versus Zoomer section is the most generous and accurate read on Gen Z I have heard from someone older than 50. Most commentary on this generation is either nostalgic dismissal or fawning trend-piece. Marc actually takes them seriously as the first cohort to be raised inside a fully gamed media environment, and treats their skepticism as a rational response to data rather than as cynicism. If you are hiring right now, this is the takeaway. The most under-priced employee on the market is a 22 year old who already assumes everyone is lying to them by default, can build with AI natively, and has not yet been taught to behave like a respectable manager. Hire them.

  • Jensen Huang on Nvidia’s Supply Chain Moat, TPU Competition, China Export Controls, and Why Nvidia Will Not Become a Cloud (Dwarkesh Podcast Summary)

    TLDW (Too Long, Didn’t Watch)

    Jensen Huang sat down with Dwarkesh Patel for over 90 minutes covering Nvidia’s supply chain dominance, the TPU threat, why Nvidia will not become a hyperscaler, whether the US should sell AI chips to China, and why Nvidia does not pursue multiple chip architectures at once. Jensen framed Nvidia’s entire business as transforming “electrons into tokens” and argued that Nvidia’s real moat is not any single technology but the full stack ecosystem it has built over two decades. He was blunt about his regret over not investing in Anthropic and OpenAI earlier, passionate about keeping the American tech stack dominant worldwide, and dismissive of the idea that China’s chip industry can be meaningfully contained through export controls.

    Key Takeaways

    1. Nvidia’s moat is the ecosystem, not the chip. Jensen repeatedly emphasized that Nvidia’s competitive advantage comes from CUDA, its massive installed base, its deep partnerships across the entire supply chain, and the fact that it operates in every cloud. The moat is not a single product but an interlocking system that took 20+ years to build.

    2. Supply chain bottlenecks are temporary, energy bottlenecks are not. Jensen argued that CoWoS packaging, HBM memory, EUV capacity, and logic fabrication bottlenecks can all be resolved in two to three years with the right demand signal. The real constraint on AI scaling is energy policy, which takes far longer to fix.

    3. TPUs and ASICs are not an existential threat to Nvidia. Jensen was emphatic that no competitor has demonstrated better price-performance or performance-per-watt than Nvidia, and challenged TPU and Trainium to prove otherwise on public benchmarks like InferenceMAX and MLPerf. He described Anthropic as a “unique instance, not a trend” for TPU adoption.

    4. Jensen regrets not investing in Anthropic and OpenAI earlier. He admitted he did not deeply internalize how much capital AI labs needed and that traditional VC funding was not sufficient for companies at that scale. He described this as a clear miss, though he said Nvidia was not in a position to make multi-billion dollar investments at the time.

    5. Nvidia will not become a hyperscaler. Jensen’s philosophy is “do as much as needed, as little as possible.” Building cloud infrastructure is something other companies can do, so Nvidia supports neoclouds like CoreWeave, Nebius, and Nscale instead of competing with them. Nvidia invests in ecosystem partners rather than vertically integrating into cloud services.

    6. Jensen is strongly against US chip export controls on China. This was the longest and most heated segment of the interview. Jensen argued that China already has abundant compute, energy, and AI researchers, and that export controls have accelerated China’s domestic chip industry while causing the US to concede the world’s second-largest technology market. He compared the situation to how US telecom policy allowed Huawei to dominate global telecommunications.

    7. AI will cause software tool usage to skyrocket, not collapse. Jensen pushed back on the narrative that AI will commoditize software companies. He argued that agents will use existing tools at massive scale, causing the number of instances of products like Excel, Synopsys Design Compiler, and other enterprise tools to grow exponentially.

    8. Nvidia does not pick winners among AI labs. Jensen explained that Nvidia invests across multiple foundation model companies simultaneously and refuses to favor any single one. He cited his own company’s unlikely survival story as the reason for this humility: Nvidia’s original graphics architecture was “precisely wrong” and would have been counted out by anyone picking winners.

    9. Nvidia added Groq for premium token economics. Nvidia recently acquired Groq and is folding it into the CUDA ecosystem because the market is now segmenting into different token tiers. Some customers will pay premium prices for faster response times even at lower throughput, creating a new segment of the inference market.

    10. Without AI, Nvidia would still be very large. Jensen was clear that accelerated computing, not AI specifically, is the foundational mission of the company. Molecular dynamics, quantum chemistry, computational lithography, data processing, and physics simulation all benefit from GPU acceleration regardless of deep learning.

    Detailed Summary

    Nvidia’s Real Business: Electrons to Tokens

    Jensen opened the conversation by reframing Nvidia’s entire value proposition. When Dwarkesh suggested that Nvidia is fundamentally a software company that sends a GDS2 file to TSMC for manufacturing, Jensen pushed back hard. He described Nvidia’s job as transforming electrons into tokens, with everything in between representing an “incredible journey” of artistry, engineering, science, and invention. He said the transformation is far from deeply understood and the journey is far from over, making commoditization unlikely.

    Jensen described Nvidia as operating a philosophy of doing “as much as necessary and as little as possible.” Whatever Nvidia does not need to do itself, it partners with someone else and makes it part of the broader ecosystem. This is why Nvidia has what Jensen called probably the largest ecosystem of partners in the industry, spanning the full supply chain upstream and downstream, application developers, model makers, and all five layers of the AI stack.

    On the question of whether AI will commoditize software companies, Jensen offered a contrarian take. He argued that agents are going to use software tools at unprecedented scale, meaning the number of instances of products like Excel, Cadence design tools, and Synopsys compilers will skyrocket. Today the bottleneck is the number of human engineers. Tomorrow, those engineers will be supported by swarms of agents exploring design spaces and using the same tools humans use today. Jensen said the reason this has not happened yet is simply that the agents are not good enough at using tools. That will change.

    The Supply Chain Moat

    Dwarkesh pressed Jensen on Nvidia’s reported $100 billion (and potentially $250 billion) in purchase commitments with foundries, memory manufacturers, and packaging companies. The question was whether Nvidia’s real moat for the next few years is simply locking up scarce upstream components so that no competitor can get the memory and logic they need to build alternative accelerators.

    Jensen confirmed this is a significant advantage but framed it differently. He said Nvidia has made enormous explicit and implicit commitments upstream. The implicit commitments matter just as much: Jensen personally meets with CEOs across the supply chain to explain the scale of the coming AI industry, convince them to invest in capacity, and assure them that Nvidia’s downstream demand is large enough to justify that investment. Nvidia’s GTC conference serves this purpose too, bringing the entire ecosystem together so upstream suppliers can see downstream demand and vice versa.

    Jensen described a process of systematically “prefetching bottlenecks” years in advance. CoWoS advanced packaging was a major bottleneck two years ago, but Nvidia swarmed it with repeated doubling of capacity until TSMC recognized it as mainstream computing technology rather than a specialty product. More recently, Nvidia has invested in the silicon photonics ecosystem through partnerships with Lumentum and Coherent, invented new packaging technologies, licensed patents to keep the supply chain open, and even invested in new testing equipment like double-sided probing.

    When Dwarkesh asked about the ultimate physical bottlenecks, Jensen surprised him. The hardest bottleneck to solve is not CoWoS or HBM or EUV machines. It is plumbers and electricians needed to build data centers. Jensen used this as a launching point to criticize “doomers” who discourage people from pursuing careers in software engineering or radiology, arguing that scaring people out of these professions creates the real bottlenecks.

    On EUV and logic scaling specifically, Jensen was optimistic. He said no supply chain bottleneck lasts longer than two to three years. Once you can build one of something, you can build ten, and once you can build ten, you can build a million. The key is a clear demand signal. If TSMC is convinced of the demand, ASML will produce enough EUV machines. Meanwhile, Nvidia continues to improve computing efficiency by 10x to 50x per generation through architecture, algorithms, and system design.

    The TPU Question

    Dwarkesh pushed hard on whether Google’s TPUs represent a real threat, noting that two of the top three AI models (Claude and Gemini) were trained on TPUs. Jensen drew a sharp distinction between what Nvidia builds and what a TPU is. Nvidia builds accelerated computing, which serves molecular dynamics, quantum chromodynamics, data processing, fluid dynamics, particle physics, and AI. A TPU is a tensor processing unit optimized for matrix multiplies. Nvidia’s market reach is far greater than any TPU or ASIC can possibly have.

    Jensen emphasized programmability as Nvidia’s core architectural advantage. If you want to invent a new attention mechanism, build a hybrid SSM model, fuse diffusion and autoregressive techniques, or disaggregate computation in a novel way, you need a generally programmable architecture. The only way to achieve 10x or 100x performance leaps (versus the roughly 25% per year from Moore’s Law) is to fundamentally change the algorithm, and that requires the flexibility CUDA provides.

    On the specific question of whether hyperscalers with huge engineering teams can simply write their own kernels and bypass CUDA, Jensen acknowledged they do write custom kernels but argued that Nvidia’s engineers still routinely deliver 2x to 3x speedups when they optimize a partner’s stack. He described Nvidia’s GPUs as “F1 racers” that anyone can drive at 100 mph, but extracting peak performance requires deep architectural expertise. Nvidia uses AI itself to generate many of its optimized kernels.

    Jensen was particularly blunt about public benchmarks. He pointed to Dylan Patel’s InferenceMAX benchmark and said neither TPU nor Trainium has been willing to demonstrate their claimed performance advantages on it. He said Nvidia’s performance-per-TCO is the best in the world, “bar none,” and challenged anyone to prove otherwise.

    Regarding Anthropic’s multi-gigawatt deal with Broadcom and Google for TPUs, Jensen called it “a unique instance, not a trend.” He said without Anthropic, there would be essentially no TPU growth and no Trainium growth. He traced this back to his own mistake: when Anthropic and OpenAI needed multi-billion dollar investments from their compute suppliers to get off the ground, Nvidia was not in a position to provide that capital. Google and AWS were, and in return, Anthropic committed to using their compute.

    Nvidia’s Investment Strategy and Regrets

    Jensen was unusually candid about his regret over not investing in foundation model companies earlier. He said he did not deeply internalize how different AI labs were from typical startups. A traditional VC would never put $5 to $10 billion into a single AI lab, but that was exactly what companies like OpenAI and Anthropic needed. By the time Jensen understood this, Nvidia was not in a financial or cultural position to make those kinds of investments.

    Now, Nvidia has invested approximately $30 billion in OpenAI and $10 billion in Anthropic. Jensen said he is delighted to support both and considers their existence essential for the world. But he acknowledged that these investments came at much higher valuations than would have been possible years earlier.

    Jensen explained Nvidia’s broader investment philosophy: support everyone, do not pick winners. He invests in one foundation model company, he invests in all of them. This comes from hard-won humility. When Nvidia started, there were 60 3D graphics companies. Nvidia’s original architecture was “precisely wrong” and the company would have been at the top of most lists to fail. Jensen said he has enough humility from that experience to know that you cannot predict which AI company will ultimately succeed.

    Why Nvidia Will Not Become a Hyperscaler

    Dwarkesh pointed out that Nvidia has the cash to build and operate its own cloud infrastructure, bypassing the middleman ecosystem that converts CapEx into OpEx for AI labs. Jensen rejected this path based on his core operating philosophy.

    If Nvidia did not build its computing platform, NVLink, and the CUDA ecosystem, nobody else would have done it. He is “completely certain” of that. These are things Nvidia must do. But the world has lots of clouds. If Nvidia did not build a cloud, someone else would show up. So the answer is to support the ecosystem instead: invest in CoreWeave, Nscale, Nebius, and others to help them exist and scale, rather than competing with them.

    Jensen was clear that Nvidia is not trying to be in the financing business either. When OpenAI needed a $30 billion investment before its IPO, Nvidia stepped up because OpenAI needed it and Nvidia deeply believed in the company. But these are targeted ecosystem investments, not a strategic pivot into cloud services.

    On GPU allocation during shortages, Jensen pushed back on the narrative that Nvidia strategically “fractures” the market by giving allocations to smaller neoclouds. He said the process is straightforward: you forecast demand, you place a purchase order, and it is first in, first out. Nvidia never changes prices based on demand. Jensen said he prefers to be dependable and serve as the foundation of the industry rather than extracting maximum short-term value.

    The China Debate

    The longest and most heated section of the interview was Jensen’s case against US chip export controls on China. This was a genuine debate, with Dwarkesh pushing the national security argument and Jensen pushing back forcefully.

    Jensen’s core argument rested on several pillars. First, China already has abundant compute. They manufacture 60% or more of the world’s mainstream chips, have massive energy infrastructure (including empty data centers with full power), and employ roughly 50% of the world’s AI researchers. The threshold of compute needed to build models like Anthropic’s Mythos has already been reached and exceeded by China’s existing infrastructure.

    Second, export controls have backfired. They accelerated China’s domestic chip industry, forced their AI ecosystem to optimize for internal architectures instead of the American tech stack, and caused the United States to concede the second-largest technology market in the world. Jensen compared this directly to how US telecom policy allowed Huawei to dominate global telecommunications infrastructure.

    Third, Jensen argued that AI is a five-layer stack (energy, chips, computing platform, models, applications) and the US needs to win at every layer. Fixating on one layer (models) at the expense of another layer (chips) is counterproductive. If Chinese open source AI models end up optimized for non-American hardware and that stack gets exported to the global south, the Middle East, Africa, and Southeast Asia, the US will have lost something far more valuable than whatever marginal compute advantage the export controls provided.

    Dwarkesh countered with the Mythos example: Anthropic’s new model found thousands of high-severity zero-day vulnerabilities across every major operating system and browser, including one that had existed in OpenBSD for 27 years. If China had enough compute to train and deploy a model like Mythos at scale before the US could prepare, the cyber-offensive capabilities would be devastating.

    Jensen’s response was direct. Mythos was trained on “fairly mundane capacity” that is already abundantly available in China. The amount of compute is not the bottleneck for that kind of breakthrough. Great computer science is, and China has no shortage of brilliant AI researchers. He pointed to DeepSeek as evidence: most advances in AI come from algorithmic innovation, not raw hardware. If China’s researchers can achieve breakthroughs like DeepSeek with limited hardware, imagine what they could do with more.

    Jensen also argued for dialogue over confrontation. He said it is essential that American and Chinese AI researchers are talking to each other, and that both countries agree on what AI should not be used for. The idea that you can prevent AI risks by cutting off chip sales, when the real advances come from algorithms and computer science, reflects a fundamental misunderstanding of how AI progress works.

    The debate ended without resolution, but Jensen’s final point was sharp: “I’m not talking to somebody who woke up a loser. That loser attitude, that loser premise, makes no sense to me.”

    Why Not Multiple Chip Architectures?

    Near the end of the interview, Dwarkesh asked why Nvidia does not run multiple parallel chip projects with different architectures, like a Cerebras-style wafer-scale design or a Dojo-style huge package, or even one without CUDA.

    Jensen’s answer was simple: “We don’t have a better idea.” Nvidia simulates all of these alternative approaches in its internal simulators and they are provably worse. The company works on exactly the projects it wants to work on. If the workload were to change dramatically (not just the algorithms, but the actual market shape), Nvidia might add other accelerators.

    In fact, Nvidia recently did exactly this by acquiring Groq. The inference market is now segmenting into different tiers. Some customers will pay premium prices for extremely fast response times even if throughput is lower. This creates a new “high ASP token” segment that justifies a different point on the performance curve. But Jensen was clear: if he had more money, he would put it all behind Nvidia’s existing architecture, not diversify into alternatives.

    Nvidia Without AI

    Jensen closed by saying that even if the deep learning revolution had never happened, Nvidia would be “very, very large.” The premise of the company has always been that general-purpose computing cannot scale indefinitely and that domain-specific acceleration is the way forward. Molecular dynamics, seismic processing, image processing, computational lithography, quantum chemistry, and data processing all benefit from GPU acceleration regardless of AI. Jensen said the fundamental promise of accelerated computing has not changed “not even a little bit.”

    Thoughts

    This interview is one of the most revealing Jensen Huang conversations in years, partly because Dwarkesh actually pushes back instead of lobbing softballs. A few things stand out.

    The Anthropic regret is real and significant. Jensen is essentially admitting that Nvidia’s biggest strategic miss of the AI era was not understanding that foundation model companies needed supplier-level capital commitments, not VC funding. The fact that Google and AWS used compute investments to lock in Anthropic’s architecture choices has had downstream consequences that Nvidia is still working to unwind. When Jensen says Anthropic is “a unique instance, not a trend” for TPU adoption, he is simultaneously downplaying the threat and revealing exactly how seriously he takes it.

    The China debate is the highlight. Jensen’s argument is more nuanced than it first appears. He is not saying “sell China everything.” He is saying the current binary approach of near-total restriction has backfired by accelerating China’s domestic chip industry and pushing the Chinese AI ecosystem away from the American tech stack. His comparison to the US telecom industry losing global market share to Huawei is pointed and historically grounded. Whether you agree with his conclusion or not, the framing of AI as a five-layer stack where the US needs to compete at every layer is a useful mental model.

    The “electrons to tokens” framing is Jensen at his best. It is a simple metaphor that captures something genuinely complex about where value is created in the AI supply chain. And his insistence that the transformation is “far from deeply understood” is a subtle way of arguing that Nvidia’s competitive position will be durable because the problem space is not close to being solved.

    The Groq acquisition reveal is interesting for what it signals about the inference market. If Nvidia is creating a separate product tier for premium-priced, low-latency tokens, it suggests the company sees inference economics fragmenting significantly. This aligns with the broader trend of AI becoming an enterprise product where different customers have wildly different willingness to pay based on how they use tokens.

    Finally, Jensen’s refusal to diversify chip architectures is a bold bet. “We simulate it all in our simulator, provably worse” is an incredibly confident statement. History is full of companies that were right until they were not. But Nvidia’s track record of 50x generation-over-generation improvements through co-design across processors, fabric, libraries, and algorithms is hard to argue with. The question is whether the current paradigm of transformer-based models on GPU clusters represents a local or global optimum for AI compute.