PJFP.com

Pursuit of Joy, Fulfillment, and Purpose

Tag: databricks

  • Inkling: Thinking Machines Lab Releases Its First Open-Weights Model, a 975B Multimodal Mixture-of-Experts With Controllable Thinking Effort That Can Fine-Tune Itself on Tinker

    Thinking Machines Lab, the AI startup founded by former OpenAI CTO Mira Murati, has released Inkling, its first open-weights model trained from scratch. Inkling is a 975 billion parameter Mixture-of-Experts transformer (41B active) with a context window of up to 1 million tokens, native multimodal reasoning over text, images, and audio, and a dial for controllable thinking effort. The lab is explicit that Inkling is not the strongest model in the world. It is pitched as something arguably more useful: a broad, balanced, customizable foundation you can fine-tune on Tinker, with the full weights on Hugging Face. The announcement even includes a demo where Inkling fine-tunes itself and swaps in its own new weights.

    TLDR

    Thinking Machines Lab released Inkling, a 975B-total, 41B-active Mixture-of-Experts model pretrained on 45 trillion tokens of text, images, audio, and video, alongside a preview of Inkling-Small (276B total, 12B active). The release covers the model’s generalist benchmark profile across reasoning, agentic coding, tool use, vision, and audio; a controllable thinking effort setting that lets developers trade performance against tokens (matching Nemotron 3 Ultra on Terminal Bench 2.1 at roughly a third of the tokens); an encoder-free multimodal architecture using dMel spectrograms and hMLP image patches; a training recipe combining Muon and Adam with weight decay coupled to the learning rate; RL scaled past 30 million rollouts with log-linearly improving reasoning and an emergent compression of the chain of thought; an epistemics push covering calibration, forecasting (where it beats several frontier models), abstention, and censorship resistance; the strongest FORTRESS adversarial safety score among compared open-weights models; a headline-grabbing demo of the model fine-tuning itself into a lipogram assistant via Tinker; and day-one availability on Tinker (at a 50% discount), Hugging Face, and inference partners including Together, Fireworks, Modal, Databricks, Baseten, vLLM, SGLang, and llama.cpp.

    Thoughts

    The most striking thing about this launch is its honesty. Nearly every frontier release leads with a claim to be the best at something, and the fine print walks it back. Thinking Machines Lab says plainly that Inkling is not the strongest model available, open or closed, and then makes the case that “strongest” is the wrong axis for most real buyers. If you are going to run a model millions of times inside a product, what you care about is the cost curve, the adaptability, and whether you can shape it to your workflow. That framing conveniently matches their business (Tinker sells fine-tuning), but it also matches how production AI actually gets deployed, where cost and latency are binding constraints and a benchmark crown is trivia.

    The self-fine-tuning demo deserves more attention than it will probably get. Asked to become a lipogram assistant that never uses the letter “e” (a behavior prompting alone cannot reliably produce), Inkling wrote its own training objective and scoring function, generated its own synthetic data, launched the run on Tinker, evaluated the result against its base self, and then staged a weight swap so the improved checkpoint took over the session. That is a closed loop of specify, train, evaluate, and self-update, packaged as a cute product demo. The loop is the primitive behind every serious conversation about recursive self-improvement, and here it is running as a marketing asset with a 27 minute wall clock. The gap between “toy objective” and “economically meaningful objective” is now a question of reward design, not plumbing.

    Controllable thinking effort is the feature I expect developers to care about most. Instead of publishing a single score, TML publishes a curve: sweep the effort setting from 0.2 to 0.99 and watch performance trade against generated tokens. Inkling reportedly matches Nemotron 3 Ultra on Terminal Bench 2.1 while spending about a third of the tokens. Benchmarks reported as single points hide exactly this, and a model that reaches a target score cheaply beats a model that scores two points higher at triple the cost in any high-volume workload. Expect effort curves to become standard marketing for open models, the way context length became standard a couple of years ago.

    The epistemics section is quietly the most differentiated part of the release. TML trained calibration directly, running RL against proper scoring rules on resolved real-world questions, and pairing a rubric grader with a claims grader that does agentic web search to verify each factual assertion. The result is a model that beats GPT-5.5 and Claude Opus 4.8 on ForecastBench without search and holds its own on Prophet Arena. A model that knows when to say “I don’t know” is more useful across messy real-world domains than one that confabulates confidently, and it is notable that a lab whose stated mission is extending human will and judgment treats calibrated uncertainty as a first-class training target rather than a safety afterthought. The censorship-resistance training, validated on Cognition’s Propaganda and Censorship Eval, extends the same idea: trustworthiness as a capability you train, not a policy you bolt on.

    Finally, the open-weights safety tension is handled with unusual candor. Inkling posts the strongest adversarial FORTRESS score among the open models compared while keeping benign over-refusal low, and it was tested externally for CBRN, cyber, and loss-of-control capabilities. But everyone in this space knows fine-tuning can strip safety behavior from open weights, and TML ships a fine-tuning platform for this exact model. Their acknowledgment that they are actively studying how safety behavior survives fine-tuning on Tinker is the right thing to say, and it is also the open question that will define whether “safe open weights” is a coherent category at all.

    Key Takeaways

    • Inkling is Thinking Machines Lab’s first from-scratch, open-weights model: a Mixture-of-Experts transformer with 975B total parameters, 41B active, and a context window up to 1M tokens.
    • It was pretrained on 45 trillion tokens spanning text, images, audio, and video, and reasons natively over text, images, and audio without separate encoders.
    • A preview of Inkling-Small ships alongside it: a 276B-parameter MoE with just 12B active parameters that matches or beats its larger sibling on several benchmarks thanks to an improved pretraining recipe.
    • TML explicitly positions Inkling as a base for customization rather than the strongest overall model, leaning on multimodality, efficient thinking, and Tinker fine-tuning as the differentiators.
    • The launch demo shows Inkling fine-tuning itself: it wrote its own training objective and data, ran the job through the Tinker API, evaluated the result, and hot-swapped to its own new weights inside the OpenCode harness.
    • The self-fine-tuning target was a lipogram assistant that never uses the letter “e,” a behavior chosen precisely because prompting alone cannot reliably achieve it; the full loop completed in about 27 minutes.
    • Controllable thinking effort is a core feature: a setting swept from 0.2 to 0.99 traces a full performance-versus-tokens curve instead of a single benchmark point.
    • On Terminal Bench 2.1, Inkling matches Nemotron 3 Ultra’s score at roughly one third of the generated tokens, the release’s flagship efficiency claim.
    • Inkling was trained to run inside a variety of coding and agent harnesses, with tool sets and schemas randomized during training to reduce sensitivity to any particular harness.
    • On Design Arena’s blinded human-evaluated Agentic Web Dev leaderboard, Inkling scores 1257, among the strongest open-weights models and tied with Claude Opus 4.6.
    • Headline benchmark scores at effort 0.99 include SWEBench Verified 77.6%, SWEBench Pro Public 54.3%, Terminal Bench 2.1 63.8%, GPQA Diamond 87.2%, AIME 2026 97.1%, and HLE 29.7% text-only (46.0% with tools).
    • Agentic and general scores include MCP Atlas 74.1%, Tau 3 Banking 23.7%, and BrowseComp 77.1% with context management.
    • Vision results are strong for an open model: MMMU Pro 73.5%, CharXiv RQ 78.1%, rising to 82.0% when the model uses a Python tool for zooming and cropping during visual reasoning.
    • Audio results place it among the strongest open-weights audio models: VoiceBench 91.4%, MMAU 77.2%, and Audio MC 56.6%, well ahead of Qwen3-Omni and Nemotron Nano-Omni on the last.
    • The multimodal stack is encoder-free: audio enters as discrete dMel spectrograms and images as 40×40 pixel patches through a four-layer hMLP, both passed through a lightweight embedding layer and processed jointly with text tokens.
    • The MoE design largely follows DeepSeek-V3: 256 routed experts plus 2 shared experts per layer, 6 routed experts active per token, with a sigmoid router and auxiliary-loss-free load balancing.
    • Attention interleaves sliding-window and global layers at a 5:1 ratio with 8 KV heads, and uses a learned relative positional embedding instead of RoPE, which TML found extrapolates better to long sequences.
    • Short convolutions are applied after the key and value projections and on the attention and MLP residual branch outputs, an unusual architectural touch aimed at efficiency and long-context performance.
    • Training used a hybrid optimizer strategy, Muon for large matrix weights and Adam for everything else, with weight decay coupled to the square of the learning rate to keep weight magnitudes stable.
    • Post-training was bootstrapped with a small SFT phase on synthetic data generated by open-weights models including Kimi K2.5, with the large majority of compute spent on large-scale RL.
    • RL was scaled past 30 million rollouts across two long continuous runs, with reasoning performance on a held-out aggregate (AIME, HLE, GPQA, and others) improving log-linearly throughout.
    • Effort control was trained by varying the system message and per-token cost across rollouts, teaching the model to modulate its own thinking budget.
    • An emergent effect appeared during RL: the chain of thought compressed over training, dropping articles and connectives into a telegraphic style, driven purely by efficiency pressure rather than any targeted reward.
    • Inkling was TML’s first major training effort and ran on NVIDIA GB300 NVL72 systems; the lab says future models will push compute scale further across pretraining and RL.
    • Calibration was trained directly with RL against proper scoring rules on a large corpus of resolved real-world questions, treating well-placed confidence as a capability rather than a byproduct.
    • On ForecastBench without search, Inkling’s Brier Index of 61.1 beats GPT-5.5 (59.1) and Claude Opus 4.8 (54.6), and it stays competitive with search enabled and on Prophet Arena.
    • Instruction following was trained with two automated graders working together: a rubric grader scoring against a checklist and a claims grader that verifies each factual claim via agentic web search, improving helpfulness and reducing hallucination simultaneously.
    • Abstention-aware rewards on short-form factual QA taught the model to answer when confident and hedge or decline when not, with some prompts explicitly forcing or forbidding hedging so the user’s preference wins.
    • Inkling was trained to answer directly on topics subject to censorship, and Cognition’s Propaganda and Censorship Eval found strong censorship non-compliance.
    • On FORTRESS, Inkling posts the strongest adversarial refusal score (78.0%) of any compared open-weights model while keeping benign compliance high (95.9%), and scores 98.6% on StrongREJECT.
    • Safety testing covered CBRN, cyber, and loss-of-control capabilities plus human-AI threat vectors like sycophancy, vulnerable users, and manipulation, verified by commissioned external testers.
    • Inkling is available for fine-tuning on Tinker today with 64K and 256K context options at a 50% limited-time discount, plus a free Inkling Playground chat interface in the Tinker console.
    • Full weights are on Hugging Face, including an NVFP4 checkpoint for efficient inference on NVIDIA Blackwell, with API availability via Together, Fireworks, Modal, Databricks, and Baseten and inference support in SGLang, vLLM, TokenSpeed, and llama.cpp.
    • TML frames Inkling as the first in a family and as the intended background reasoning model for its previously announced real-time interaction models system.

    Detailed Summary

    What Inkling Is and Why It Exists

    Thinking Machines Lab frames its mission as building AI that extends human will and judgment, and Inkling as the logical next step after shipping the Tinker customization platform, previewing an interaction-focused AI system, and publishing research. Inkling is a Mixture-of-Experts transformer with 975B total and 41B active parameters, a context window up to 1M tokens, and pretraining on 45 trillion tokens of mixed text, image, audio, and video data. The lab is upfront that it is not the strongest model available. The pitch is breadth plus adaptability: a generalist trained across agentic, reasoning, coding, instruction-following, factuality, vision, and audio tasks rather than tuned to dominate one leaderboard, offered with full weights so people can make it their own. It launches with a preview sibling, Inkling-Small, at 276B total and 12B active parameters.

    The Self-Fine-Tuning Demo

    To demonstrate what customization means, TML asked Inkling to fine-tune itself. Running inside the OpenCode harness with access to Tinker, the model was told to become a lipogram assistant that never uses the letter “e.” Inkling drafted the plan, wrote an objective file with a scoring function (any response containing “e” scores zero), generated synthetic training data, launched a supervised fine-tuning run through the Tinker API, evaluated the checkpoint against its base self, and then staged a self-update so the supervisor relaunched the session on the new weights. The pipeline passed in about 27 minutes, and the updated model answered a test question about launching an LLM without a single “e.” It is a whimsical objective wrapped around a serious primitive: a model autonomously specifying, running, and adopting its own weight updates.

    Agentic Coding and Tool Use

    TML trained Inkling to operate inside many coding and agent harnesses, randomizing tool sets and schemas during training so the model does not overfit to one environment. The release showcases three demos: a one-shot job-application web app that then hosts an embedded browser-use agent operating its own interface; a nine-page, cohesively designed PDF food and travel journal produced from a single editorial prompt with web-verified details; and a server-authoritative multiplayer snake game refined over 40 iterations of feedback from GPT Codex acting as a reviewer. On benchmarks, Inkling posts 77.6% on SWEBench Verified, 54.3% on SWEBench Pro Public, and 63.8% on Terminal Bench 2.1, competitive within the open-weights field, and 1257 on Design Arena’s human-judged web dev leaderboard, in the same band as Claude Opus 4.6.

    Controllable Thinking Effort

    Rather than reporting a single operating point, TML sweeps Inkling’s effort setting from 0.2 to 0.99 and plots score against mean generated tokens on Terminal Bench 2.1, HLE, and IFBench, with competitors shown at their default settings. The headline result is efficiency: Inkling reaches Nemotron 3 Ultra’s Terminal Bench score at roughly a third of the tokens. The argument is that cost and latency are binding constraints in production, especially for interactive collaboration, so the full cost curve, not the peak score, is what developers should evaluate. Effort can be set from within the agent harness, and the ability was trained by varying system messages and per-token costs across RL rollouts.

    Native Multimodality Without Encoders

    Inkling is designed to serve as the background reasoning model for TML’s interaction models system, which requires real-time voice and vision collaboration. The multimodal components are trained from scratch with an encoder-free architecture: audio arrives as discrete dMel spectrograms and images as 40×40 pixel patches through a four-layer hMLP, both mapped through a lightweight embedding layer and processed jointly with text. The model transcribes speech, follows spoken instructions, reasons over long recordings, and answers questions about charts and diagrams, optionally using a Python tool to zoom and crop images mid-reasoning. Scores like 91.4% on VoiceBench and 82.0% on CharXiv RQ with Python place it among the strongest open-weights multimodal models, though still behind Gemini 3.1 Pro.

    Epistemics: Calibration, Forecasting, and Censorship Resistance

    TML groups calibration, instruction following, and censorship resistance under the banner of epistemics. Calibration was trained with RL against proper scoring rules on resolved real-world questions, and it shows: Inkling’s ForecastBench Brier Index of 61.1 without search beats GPT-5.5 and Claude Opus 4.8, and its Prophet Arena score sits close to the frontier. Instruction following used two complementary automated graders, a rubric checklist and a claims grader that verifies factual assertions through agentic web search, so recall-spraying to hack rubrics gets penalized by the factuality check. Targeted abstention-aware QA datasets taught the model to say “I don’t know” or give hedged best guesses when appropriate, while still complying when a user demands a forced guess. Finally, the model was trained to answer directly on censorship-prone topics, with Cognition’s Propaganda and Censorship Eval finding strong non-compliance with censorship patterns.

    Safety for an Open-Weights Release

    Inkling was trained to an internal behavioral spec across all modalities and then checked by commissioned external safety testers. Evaluations covered dangerous capabilities (CBRN, cyber, loss of control) and human-AI threat vectors including sycophancy, vulnerable users, and harmful manipulation. On FORTRESS, which pairs adversarial harmful requests with benign look-alikes, Inkling posts the strongest adversarial score among the compared open models (78.0%) without collapsing on the benign side (95.9%), and it scores 98.6% on StrongREJECT. TML acknowledges the open question hanging over every open-weights release: how safety behavior holds up under fine-tuning, which it says it is actively studying on Tinker.

    Architecture and Training Recipe

    The MoE layout follows DeepSeek-V3: 256 routed experts and 2 shared experts per layer with 6 routed experts active per token, a sigmoid-based router, and auxiliary-loss-free load balancing. Attention interleaves sliding-window and global layers 5:1 with 8 KV heads, and positions are encoded with a learned relative positional embedding that TML found outperforms and out-extrapolates RoPE. Short convolutions appear after the key and value projections and on residual branch outputs. Optimization was hybrid, Muon for large matrices and Adam elsewhere, with hyperparameter schedules drawn from the lab’s modular manifolds research and weight decay coupled to the square of the learning rate to keep weight norms stable. Post-training bootstrapped from a small SFT phase on synthetic data from open models including Kimi K2.5, then spent the bulk of compute on large-scale RL. Everything ran on NVIDIA GB300 NVL72 systems.

    RL at Scale and the Emergent Compression of Thought

    TML scaled asynchronous RL past 30 million rollouts across two long continuous runs, with performance on a held-out aggregate of reasoning evals improving log-linearly the whole way. Along the way an unplanned behavior emerged: the chain of thought became progressively more concise, shedding grammatical overhead into a telegraphic style (“We need to understand” becomes “We need determine”) while remaining comprehensible and leaving final answers unaffected. No reward targeted this; token efficiency pressure alone drove the compression, echoing an observation Cognition made while training SWE-1.7. It is a vivid example of optimization discovering its own shorthand.

    Inkling-Small

    The preview of Inkling-Small is arguably the sleeper story: with 12B active parameters against Inkling’s 41B, it matches or exceeds the larger model on a surprising number of benchmarks, including GPQA Diamond (88.3% vs 87.2%), IFBench (83.4% vs 79.8%), and CharXiv RQ with Python (83.4% vs 82.0%). TML attributes this to pretraining data and recipe improvements made after the big model trained, with both models sharing the same post-training stack. The clearest gaps favoring big Inkling are factuality (SimpleQA 43.9% vs 20.9%), Terminal Bench, and Tau 3 Banking. Full weights for Inkling-Small will be released once testing finishes, and its cost and latency profile targets high-volume workloads like coding, LLM grading, and synthetic data generation.

    Availability and the Ecosystem Play

    Inkling is on Tinker today with 64K and 256K context options at a limited-time 50% discount, plus a free Inkling Playground chat interface with integrated web search in the Tinker console so developers can get a feel for the model before committing to a run. The cookbook gained native Inkling support and three new audio recipes, and a new tml-renderer handles chat templates, tool calls, reasoning content, and multimodal inputs. Deployment partnerships span Together, Fireworks, Modal, Databricks, and Baseten for APIs; RadixArk for SGLang and Miles; Inferact for vLLM; Lightseek for TokenSpeed; Unsloth for llama.cpp; and Hugging Face for transformers integration. Full weights are on Hugging Face in both the original checkpoint and an NVFP4 checkpoint for NVIDIA Blackwell inference.

    Notable Quotes

    “Our mission is to build AI that extends human will and judgment.”

    Thinking Machines Lab, opening the Inkling announcement

    The company’s north star, and the lens through which the whole release (customization, calibration, open weights) is framed.

    “Inkling is not the strongest overall model available today, open or closed. Instead, a combination of qualities makes it a good open-weights base for customization: multimodal capabilities, efficient thinking, and availability on Tinker for fine-tuning.”

    Thinking Machines Lab, positioning the release

    A rare piece of launch-day honesty from a frontier lab, and the strategic thesis of the whole release.

    “Picking the right base model to fine-tune is a qualitative judgment that combines measurable benchmarks with the unique feel of a model that comes from playing with it.”

    Thinking Machines Lab, on why the Inkling Playground exists

    An argument that vibes are data, from the lab that built a playground into a fine-tuning console.

    “Cost and latency are often binding constraints in real-world applications, and low latency in particular is crucial for enabling collaboration and improvement through iteration.”

    Thinking Machines Lab, on controllable thinking effort

    The case for evaluating models on their full effort-versus-performance curve instead of a single benchmark point.

    “A model that’s confident in every answer it gives, including when it’s missing info and confabulates, forces the user to double-check everything.”

    Thinking Machines Lab, on why calibration was a training target

    The clearest one-line justification for treating calibrated uncertainty as a capability rather than a nicety.

    “Together, the two graders improve helpfulness and reduce hallucination at the same time, rather than trading one for the other.”

    Thinking Machines Lab, on pairing a rubric grader with a web-searching claims grader

    A neat solution to rubric hacking: verify every claim with agentic search so spraying plausible facts stops paying.

    “Safety is crucial for open-weights models. We’re continuing to study safety behavior and capability uplift in customizable models, including how safety behavior is impacted by fine-tuning on Tinker.”

    Thinking Machines Lab, on the open question of fine-tunable safety

    The acknowledgment that safety trained into open weights must survive the very customization the product sells.

    “Inkling is just the start: our first release in a model family we will continue to build on.”

    Thinking Machines Lab, on the roadmap

    Together with the GB300 compute note, a clear signal that larger and stronger family members are coming.

    Read the full announcement, including the interactive demos, effort curves, and complete benchmark tables, on the Thinking Machines Lab blog.

    Related Reading

  • Thomas Laffont of Coatue on the $4 Trillion AI IPO Wave: SpaceX, Anthropic, OpenAI, and Why the New Unicorn Economy Is Healthier

    Thomas Laffont, co-founder of the $55 billion hedge fund Coatue Management, made his All-In Podcast premiere with a data-dense walk through what he calls a once-in-a-generation moment for the unicorn economy. In front of Chamath Palihapitiya, Jason Calacanis, David Sacks, and David Friedberg, he argued that a roughly $4 trillion wave of private value is about to hit the public markets, led by SpaceX, Anthropic, and OpenAI, and that the new AI-driven unicorn economy is actually healthier than the one that came before it. You can watch the full presentation and Q&A on YouTube.

    TLDW

    Laffont presents Coatue’s slide deck on the state of the unicorn economy and argues it has rebalanced after the excesses of 2021. The average unicorn is up about 70 percent since September 2024, AI keeps taking a bigger share of all fundraising, and the model has shifted from many small unicorns to fewer companies each raising far more, with funding per unicorn up roughly 5x since 2021. He introduces a “Magnificent 8” private index (SpaceX, Stripe, Anthropic, Databricks, Revolut, ByteDance, Anduril, and more) worth nearly $4 trillion that has crushed the public Mag 7, then shows that exits are finally thawing as SpaceX heads to an IPO in weeks and Anthropic confidentially files its S1. He lays out Coatue’s “CODE” framework for why SpaceX gets more valuable the more it launches, a counterintuitive finding that the odds of a 10x actually rise as companies get bigger (31 percent for $100 billion-plus centicorns), the explosive revenue ramp of OpenAI and Anthropic past Workday, ServiceNow, Adobe, Salesforce, and now the hyperscalers, a three-pillar map of where AI revenue comes from (consumer, ads, enterprise), and the AI memory thesis. The Q&A with Chamath and Calacanis digs into the power law, K-shaped outcomes, whether these valuations are disconnected from reality, the public market as the great antiseptic, and what happens when trillions in private value finally recycles back through GPs and LPs.

    Thoughts

    The most useful idea in the talk is not the $4 trillion headline, it is the cohort-health chart. Laffont splits unicorns into eras and shows that the pre-2021 cohort was healthy, roughly 80 percent had raised again or exited 20 quarters after minting, while the giant 2021 ZIRP cohort of 479 companies is stuck with under 20 percent doing either. That single comparison reframes the whole AI boom. The bullish read is that the 2024 AI cohort is small, concentrated, and cash-generative, so it looks more like the healthy pre-ZIRP group than the 2021 hangover. The bearish read is that we are watching the same movie with bigger numbers, and the test only comes when these companies face public markets. Laffont is honest that we do not yet know which cohort the AI class resembles, and that intellectual humility is what makes the deck credible rather than promotional.

    The SpaceX “CODE” framework is the sharpest analytical move of the presentation. Most people would assume a launch business gets cheaper per launch as it scales. Laffont shows the opposite, the market pays more per launch as cadence rises, and explains it as a phase change in business quality: from one-time government launch revenue, to a single recurring-revenue constellation, to multiple constellations, to a platform with optional upside in space data centers, the moon, and Mars. It is a clean way to think about any company that climbs from a project business to a platform business, and it applies far beyond rockets. The lesson for investors is that valuation can rationally expand even as unit economics look like they should compress, because the nature of the revenue underneath is changing.

    The counterintuitive 10x odds finding deserves more attention than it got in the room. Conventional wisdom says the bigger you are, the harder it is to grow, so a $100 billion company should be less likely to 10x than a $10 billion one. Coatue’s data says the reverse: centicorns have a 31 percent shot at a 10x, far higher than the 8 percent a unicorn has at becoming a decacorn. Laffont’s explanation is a filtering mechanism, every step up validates a compounding advantage and durability of earnings, so survivors are increasingly the kind of business that keeps compounding. This is essentially a quantitative restatement of quality investing, and it is the intellectual backbone of the LP strategy the besties tease out, just buy whoever reaches $100 billion and hold.

    Where the argument gets genuinely contested is valuation, and the panel does not let it slide. The pushback that “these are not fake companies” is true and important, OpenAI and Anthropic are growing faster than any software company in history, and Anthropic reportedly had a profitable month. But growth and reality do not settle the question of price when you are paying 50 to 100 times revenue for trillion-dollar private companies, as Bill Ackman pointed out earlier in the day. Laffont’s answer is the most grounded thing he says all session: the public market is the great antiseptic, it will not care about anyone’s slide deck, and he wants to see these names withstand short sellers and skeptics. That is the right posture. The deck is a thesis, not a verdict, and the verdict arrives roughly six months and one day after the IPOs, once passive flows and supply have washed through.

    The closing thread, that almost every sector is being transformed at once and we still do not have superintelligence, is the part worth sitting with. The risk in a presentation this bullish is treating the trend as destiny. The value is in the framing tools Laffont hands you, cohort health, phase-change business quality, the filtering odds, the three revenue pillars, and the antiseptic of public scrutiny. Use those to interrogate each name rather than to buy the index on faith, and the talk earns its premiere billing.

    Key Takeaways

    • Coatue Management is one of the most successful hedge funds of the last two decades with about $55 billion under management, and is raising roughly another billion dollars specifically to invest in AI.
    • The unicorn economy is up about 70 percent on average since September 2024, and the public market has made a similar move up over the same period.
    • The unicorn economy’s share of the NASDAQ rose significantly after 2015 but has plateaued in recent years, reflecting strong performance from public companies.
    • AI keeps increasing its wallet share of all venture fundraising, multiple years in a row now.
    • The composition of funding has changed. The unicorn “factory” peaked in the ZIRP era of 2021 and has normalized at a much lower level since.
    • Funding per unicorn has increased roughly 5x since 2021. There are fewer unicorns, and each one is raising more.
    • Cohort health, pre-ZIRP group: of about 73 unicorns, 20 quarters after minting roughly 80 percent had either raised a new round or exited, which is healthy.
    • Cohort health, 2021 group: of about 479 unicorns, 20 quarters in, fewer than 20 percent had exited or raised again. Far larger cohort, far worse outcomes.
    • The open question is which cohort the new 2024 AI cohort will resemble.
    • Funding is concentrating: the top 10 companies capture a large share, and it is a small number of AI companies, not all of them, with Anthropic and OpenAI raising massive rounds.
    • Laffont proposes a “Magnificent 8” private index: SpaceX, Stripe, Anthropic, Databricks, Revolut, ByteDance, Anduril, and more, spanning internet, AI, fintech, and space tech.
    • That private index represents almost $4 trillion of value and has crushed the traditional public Mag 7, with almost every name outperforming.
    • Exits are thawing. 2026 is on a good trend for cash returned versus consumed, not quite 2021 levels, with half a year still to go.
    • That trend does not yet include three imminent liquidity events: SpaceX (IPO expected in weeks) and Anthropic (confidentially filed its S1), whose combined value could exceed the prior decade of exits combined.
    • The ecosystem is far more balanced than when Laffont first presented at the 2024 All-In Summit, when it was consuming much more cash than it returned.
    • OpenAI and Anthropic revenue growth is unlike anything previously seen. Starting from January 2025, they passed Workday, then ServiceNow, then Adobe, then Salesforce, and are now bigger than Google Cloud and Azure.
    • On current forecasts, that revenue could pass AWS by the end of the year and exceed all of Microsoft by 2028.
    • Hyperscalers are not sitting still. The largest companies in the world are funding the disruption, investing unprecedented sums to enable the ChatGPT moment.
    • The SpaceX “CODE” framework: the number one driver correlated to SpaceX’s valuation is cadence of launches, and valuation per launch rises as launches increase.
    • Why per-launch value rises: business quality improves through phases, pre-constellation (one-time government revenue), initial ramp (one recurring-revenue constellation), scale (multiple constellations), and platform (space data centers, moon and Mars optionality).
    • Anthropic in particular is scaling like no company seen across the PC, internet, or mobile eras.
    • Counterintuitive 10x odds: a unicorn has about an 8 percent chance of becoming a decacorn, a decacorn has 8 to 13 percent odds of reaching $100 billion, but a centicorn ($100 billion-plus) has a 31 percent chance of a 10x.
    • Value creation has accelerated. It typically takes years to go from $500 billion to $1 trillion in market cap, yet recently three companies did it in one year and two did it in a matter of weeks.
    • Cerebras is the counterexample of slow success: years of dark periods and no new capital developing its technology, then a massive OpenAI contract that quintupled the company’s value ahead of its IPO.
    • Semiconductors are on a generational run, with the sector dramatically outperforming the index since the 2024 All-In Summit.
    • AI memory thesis: the more an AI system knows about you, the more useful it is, so memory per user could quintuple, which helps explain recent moves in memory companies.
    • Where the revenue is: the AI ecosystem is roughly $140 billion today, about $300 billion this year, and is expected to double in 2027.
    • Three revenue pillars: consumer (subscribers times ARPU), ads (about a quarter of Meta and Google ads are AI-enabled today, heading toward 100 percent and roughly $150 billion), and enterprise (tools like Claude Code and Codex inside businesses).
    • Disruption is hitting every sector: software, telco (Starlink-powered global phone calls), semis, energy (data centers reshaping Pennsylvania’s grid), auto (Ferrari’s electric and autonomous stumble), and consumer (GLP-1s reshaping food, alcohol, and wellness).
    • Final takeaways: the new unicorn economy is healthier thanks to AI, winners are compounding faster so the cost of not owning a winner is higher than ever, disruption is everywhere, and we do not even have superintelligence yet.
    • In the Q&A, both Anthropic and OpenAI publicly say they want to be public, and big outcomes now look likely to become liquid within roughly a 12-month window.
    • The valuation pushback: these are not fake companies, they generate substantial revenue at scale and grow faster than anything before, and Anthropic reportedly even had a profitable month.
    • The public market is framed as the great equalizer and antiseptic, but with passive buying the true price discovery may not land on day one, more like six months and a day after listing.
    • A floated LP strategy: wait for whoever reaches $100 billion and concentrate capital there as the least brittle, quickest-return bet, tempered by the warning that valuations are disconnecting from any historical metric (50x to 100x revenue).
    • An open risk: with so much capital, OpenAI and Anthropic could rationally start a price war, the way ride-sharing and food-delivery players once did, though heavy infrastructure spend complicates it.

    Detailed Summary

    The unicorn economy has rebalanced after 2021

    Laffont opens by reframing a market many assume is frothy. The average unicorn is up about 70 percent since September 2024, and the public market has tracked a similar climb, so private and public value are moving together rather than diverging. The unicorn economy’s share of the NASDAQ rose sharply after 2015 and then plateaued, which he reads as a sign of how strong public companies have become. Underneath the headline, the structure of funding has changed. The 2021 ZIRP era was a unicorn factory that minted enormous numbers of companies, and that machine has since normalized to a much lower level. The result is a barbell: fewer new unicorns, but each raising far more, with funding per unicorn up roughly 5x since 2021. AI sits at the center of this, taking a steadily larger share of all venture dollars for several years running.

    Cohort health is the real story

    The deck’s most important slide measures the health of the ecosystem by cohort. The pre-ZIRP cohort, about 73 unicorns, looks healthy: 20 quarters after becoming unicorns, roughly 80 percent had either raised a new round or exited. The 2021 cohort tells the opposite story. It is enormous, about 479 unicorns, and 20 quarters in, fewer than 20 percent had raised again or exited. That contrast sets up the central question of the talk. A new 2024 cohort of AI companies is forming, and no one yet knows whether it will resemble the healthy pre-ZIRP group or the bloated, stuck 2021 group. Laffont’s framing leans optimistic because the AI cohort is small and concentrated, but he is careful not to declare the answer.

    The Magnificent 8 and a $4 trillion private index

    Funding is not just flowing to AI, it is flowing to a handful of AI names, with the top 10 capturing a large share and Anthropic and OpenAI raising the biggest rounds. From this concentration Laffont builds a private index he half-jokingly calls the Magnificent 8, a number he expects to shrink as companies go public. The members span sectors: SpaceX, Stripe, Anthropic, Databricks, Revolut, ByteDance, and Anduril, covering internet, AI, fintech, and space tech. He says he would be comfortable owning that index for the next decade-plus. Collectively it represents almost $4 trillion of value and has outperformed the public Mag 7, with nearly every constituent beating that benchmark.

    Exits are thawing and a wall of liquidity is coming

    One of Laffont’s recurring concerns at past summits has been balance: the unicorn economy is great at consuming cash, but a healthy ecosystem must also return it. On that score 2026 is trending well, not quite 2021, but solid with half a year left. Crucially, that figure does not yet include three imminent events. SpaceX is expected to go public within weeks, and Anthropic confidentially filed its S1 the day of the talk. Adding those up, just a few companies could deliver more liquidity than the prior ten years combined. The takeaway is that the ecosystem that was dangerously out of balance in 2024 is now meaningfully more balanced, and improving.

    The revenue ramp past the hyperscalers

    The growth rates of OpenAI and Anthropic, Laffont argues, are unlike anything previously seen. Charting from January 2025, the leading AI labs passed Workday, then ServiceNow, then Adobe by year end, then Salesforce by January, and are now bigger than Google Cloud and Azure. On forecast, that revenue could surpass AWS by the end of the year and exceed all of Microsoft by 2028. He stresses that the hyperscalers are not passive bystanders, they are actively funding the disruption, pouring unprecedented capital into enabling the change that began with the ChatGPT moment.

    The SpaceX CODE framework

    Laffont devotes real time to how Coatue thinks about SpaceX. The single factor most correlated with SpaceX’s valuation is cadence of launches, which is intuitive for a launch business. The surprise is that valuation per launch has risen rather than fallen as cadence climbed. His explanation, the CODE framework, is that the quality of the business model improves the more SpaceX launches. In phase one, pre-constellation, you are simply proving rockets, with a few government customers and lumpy, unpredictable one-time revenue. In the initial ramp you stand up a constellation, which is an end market and a recurring-revenue business that grows with every satellite and subscriber. At scale you operate multiple constellations, and Laffont expects companies, governments, and militaries to want to own their own. Ultimately it becomes a platform, with new businesses layered on top, from space data centers to the optionality of the moon and Mars.

    Counterintuitive odds and the speed of value creation

    Coatue bucketed companies and asked the odds of a 10x within each. A unicorn has roughly an 8 percent chance of becoming a decacorn. A decacorn has 8 to 13 percent odds of reaching $100 billion. But a centicorn, $100 billion or more, has a 31 percent chance of a 10x, counting both public and private companies. The bigger you are, the better your odds, which inverts intuition. Laffont pairs this with the sheer speed of recent value creation. Going from $500 billion to $1 trillion in market cap normally takes years, yet three companies did it in a single year and two did it in a matter of weeks. He also offers Cerebras as the patient counterexample, a chip company that endured years of dark periods and no new capital before a massive OpenAI contract quintupled its value ahead of IPO, part of a broader generational run for semiconductors.

    AI memory and where the revenue actually comes from

    A throughline from the day’s other speakers is that the more an AI knows about you, the more useful it is, from your restaurant preferences to your work context. Laffont turns that into a thesis: memory per user could quintuple based on what these systems require, which helps explain recent moves in memory companies. He then tackles the most contested question, where is the revenue. He sizes the AI ecosystem at about $140 billion today, roughly $300 billion this year, and doubling in 2027, built on three pillars. Consumer is subscribers times ARPU. Ads are the pillar people forget, with about a quarter of Meta and Google ads already AI-enabled and penetration heading toward 100 percent, a roughly $150 billion opportunity. Enterprise is the breakthrough category, exemplified by tools like Claude Code and Codex operating inside businesses.

    Every sector is being transformed at once

    What makes this era different, Laffont says, is that nearly every sector is being transformed simultaneously. Software is obvious, but look at telco, where he believes Starlink will soon power a device that lets you make a phone call anywhere on earth, attacking the global telco and broadband profit pool with a better product. Compute is driving massive change in semis, data centers are reshaping the energy equation in places like Pennsylvania, and the auto business is being upended, as Ferrari’s stumble introducing electric and autonomous technology showed. In consumer, GLP-1 drugs are profoundly changing consumption of food and alcohol and the broader focus on wellness. His takeaways close the loop: the new unicorn economy is healthier thanks to AI, winners are compounding faster so the cost of missing them is higher than ever, disruption is everywhere, and superintelligence has not even arrived yet.

    The Q&A: power law, valuation, and the public market test

    Chamath and Jason Calacanis press Laffont on what this means for allocators. The recurring theme is the power law and K-shaped outcomes, with gains consolidating into a small number of companies. The positive side, Laffont notes, is that outcomes are enormous and increasingly liquid within a 12-month window, and both Anthropic and OpenAI say they want to be public. The hard part is valuation. The besties cite Bill Ackman’s framing that investors are making venture bets on trillion-dollar companies at 50 to 100 times revenue. Laffont’s pushback is that these are not fake companies, they generate substantial revenue at scale and grow faster than anything before, and Anthropic reportedly had a profitable month. But he embraces the discipline ahead: the public market is the great antiseptic and will not care about anyone’s presentation, though with heavy passive buying, true price discovery may take roughly six months and a day rather than landing on day one. Asked whether the compounding is a market inefficiency or survivor bias, he declines to over-read a small sample, noting that Anthropic before Claude Code was a completely different company than after. The conversation closes on what happens when trillions recycle from GPs to LPs, the case for simply owning whoever crosses $100 billion, the risk of everyone crowding into three names, and the possibility of an eventual OpenAI versus Anthropic price war.

    Notable Quotes

    “So we have fewer unicorns that are each raising more.”

    Thomas Laffont, summarizing how funding per unicorn has risen roughly 5x since 2021

    “The reason is that the quality of SpaceX’s business model increases the more you launch.”

    Thomas Laffont, explaining the CODE framework and why valuation per launch rises with cadence

    “The winners are compounding faster than ever, which means the costs of not being in a winner are higher than ever.”

    Thomas Laffont, on the central risk of a power-law market

    “And by the way, we don’t even have super intelligence yet.”

    Thomas Laffont, closing his takeaways on how early the transformation still is

    “These are companies generating substantial revenue at scale that are growing faster than anything we’ve ever seen.”

    Thomas Laffont, pushing back on the idea that AI valuations rest on fake companies

    “It will be the great antiseptic. It will not care about my presentation.”

    Thomas Laffont, on the public market as the ultimate test for SpaceX, OpenAI, and Anthropic

    “Anthropic pre-cloud code was a completely different company than post cloud code.”

    Thomas Laffont, on why he won’t over-read a small sample of hyper-compounders

    “The power law rules our lives. All the great gains are being consolidated into small numbers of companies.”

    An All-In host, framing the Q&A on concentration in private markets

    This is a curated set of highlights. To hear the full presentation, the slide walkthrough, and the complete Q&A with Chamath and Jason Calacanis, watch the full conversation here.

    Related Reading

    • Coatue Management. Primary source for Thomas Laffont’s firm and the technology investing strategy behind the deck.
    • The All-In Podcast. The show and summit where Laffont made this premiere presentation.
    • Power law (Wikipedia). Background on the distribution Laffont and the hosts say governs venture and public-market returns.
    • The Magnificent Seven (Wikipedia). The public-market benchmark Laffont’s private “Magnificent 8” index is measured against.
    • Cerebras Systems. The AI chipmaker Laffont cites as the slow-grind IPO that was eventually transformed by a major OpenAI contract.

  • Claude Opus 4.8 Released: Anthropic Bets on Honesty, Dynamic Workflows, Effort Control, and Cheaper Fast Mode

    Anthropic has released Claude Opus 4.8, the newest member of its flagship Opus class, available today across every surface and priced exactly like the model it replaces. The company calls it “a modest but tangible improvement” on Opus 4.7, but the framing undersells what is actually interesting here: the headline upgrade is not a benchmark number, it is honesty. Opus 4.8 is built to know when it does not know, and that single behavioral shift may matter more for real agent work than any raw capability bump.

    TLDR

    Claude Opus 4.8 is an across-the-board upgrade to Anthropic’s Opus class that ships today at the same regular price as Opus 4.7 ($5 per million input tokens, $25 per million output tokens), with the model positioned as “a more effective collaborator.” The marquee improvement is honesty: Opus 4.8 is roughly four times less likely than its predecessor to let flaws in its own code pass unremarked, and it is more willing to flag uncertainty rather than confidently claim progress on thin evidence. A pre-release alignment assessment found new highs on prosocial traits like supporting user autonomy and acting in the user’s best interest, with misaligned behavior at rates similar to Anthropic’s best-aligned model, Claude Mythos Preview. Three things launch alongside the model: dynamic workflows in Claude Code (research preview), where Claude plans work then runs hundreds of parallel subagents that run even longer and verify their own outputs before reporting back; effort control in claude.ai and Cowork, a slider for how hard Claude thinks; and a Messages API update that accepts system entries inside the messages array so developers can update instructions mid-task without breaking the prompt cache. Fast mode now runs at 2.5x speed and is three times cheaper than before ($10 / $50 per million tokens). The roadmap points to cheaper Opus-equivalent models, a higher-intelligence class above Opus, and a wider rollout of Mythos-class models gated behind stronger cyber safeguards under Project Glasswing.

    Thoughts

    The most important sentence in this announcement is not about coding scores. It is the claim that Opus 4.8 is about four times less likely than Opus 4.7 to let flaws in its own code slip by without comment. For a chat assistant, overconfidence is annoying. For an agent, it is catastrophic. The whole premise of long-running autonomous work is that you hand the model a task and walk away, which means the model’s own judgment about whether it succeeded becomes the only judgment in the loop until you come back. A model that confidently declares victory on a half-finished migration does not save you time, it costs you a debugging session plus the time you spent trusting it. Honesty, framed this way, is not a soft virtue. It is the load-bearing reliability property that makes unattended agents usable at all.

    Read the launch as a single coherent argument rather than a list of features, and the pieces lock together. Dynamic workflows let Claude plan a job and fan out hundreds of parallel subagents that, with Opus 4.8, run longer than before. Effort control lets you dial up how much the model thinks. The honesty improvement means the model checks its own work and flags what it is unsure about instead of papering over it. Put those three together and you get one product thesis: let it run longer, let it think harder, and trust it to tell you when something is wrong. The codebase-scale migration example, hundreds of thousands of lines from kickoff to merge with the existing test suite as the bar, is the proof point. None of those three capabilities is worth much alone. A model that runs for hours but lies about its results is a liability. A model that flags uncertainty but cannot sustain a long task never reaches the moment where its honesty matters. Anthropic shipped all three at once because they only pay off together.

    The economics deserve a closer look than the “same price” headline invites. Regular pricing is flat versus Opus 4.7, which is the polite way of saying you get a better model for free. The real move is fast mode: 2.5x the speed at three times cheaper than it cost on previous models, landing at $10 per million input and $50 per million output. That is Anthropic quietly attacking the latency-versus-cost tradeoff that has shaped how teams deploy frontier models. Until now, “fast” meant “expensive,” so you reserved it for interactive moments and ate the wait everywhere else. Collapsing that premium changes the default. And note the subtle token story underneath: Opus 4.8 at its default high effort spends roughly the same tokens on coding as Opus 4.7’s default while performing better, so the effort slider is not a way to bleed you dry, it is an honest exposure of the quality-cost dial that was always there implicitly.

    The Messages API change is the kind of unglamorous plumbing that practitioners will appreciate immediately. Letting system entries live inside the messages array means you can update an agent’s instructions, permissions, token budget, or environment context partway through a task without smuggling the update through a fake user turn and without blowing up your prompt cache. Anyone who has built a long-running agent has hit this wall: the world changes mid-task, the agent needs new constraints, and the only clean way to inject them previously was a cache-busting hack. This is Anthropic treating agents as first-class, stateful, long-lived processes rather than oversized chat sessions. It is a small spec change with outsized implications for how you architect an agent that runs for an hour.

    Then there is the roadmap, where the most telling line is the quietest. Anthropic says a small number of organizations are already using Claude Mythos Preview for cybersecurity work under Project Glasswing, and that models of this capability level require stronger cyber safeguards before general release. Notice that they are pinning Opus 4.8’s alignment numbers to Mythos as the benchmark for “best-aligned,” while simultaneously holding Mythos back from general availability on safety grounds. That is a deliberate signal: the next class of model is good enough that they are gating it on cyber-offense risk, not on capability. For a site about the pursuit of joy, fulfillment, and purpose through AI, this is the part worth sitting with. The frontier is increasingly defined not by what the models can do, but by what their builders decide it is responsible to ship. Honesty in the small (flagging a bad line of code) and restraint in the large (holding back a cyber-capable model) are the same instinct expressed at two different scales.

    Key Takeaways

    • Claude Opus 4.8 is now available everywhere, replacing Opus 4.7 as Anthropic’s flagship Opus-class model and positioned as “a more effective collaborator.”
    • Regular usage pricing is unchanged from Opus 4.7, holding at $5 per million input tokens and $25 per million output tokens, so the capability gains come at no added cost.
    • The single most emphasized improvement is honesty, which Anthropic treats as a core trained behavior rather than a marketing flourish.
    • Evaluations show Opus 4.8 is around four times less likely than its predecessor to let flaws in its own code pass unremarked, a direct reliability win for autonomous coding.
    • Early testers report the model is more likely to flag uncertainty about its work and less likely to make unsupported claims or jump to conclusions on thin evidence.
    • A detailed alignment assessment was run before release and concluded Opus 4.8 reaches new highs on prosocial traits like supporting user autonomy and acting in the user’s best interest.
    • Misaligned behavior such as deception or cooperation with misuse is at rates substantially lower than Opus 4.7 and similar to Anthropic’s best-aligned model, Claude Mythos Preview.
    • The full alignment assessment and pre-deployment safety tests are documented in the public Claude Opus 4.8 System Card.
    • Dynamic workflows launch as a research preview inside Claude Code, letting Claude plan the work and then run hundreds of parallel subagents in a single session.
    • With Opus 4.8, those subagents can run even longer, and Claude verifies its outputs before reporting back rather than declaring success blindly.
    • Anthropic’s flagship example for dynamic workflows is a codebase-scale migration across hundreds of thousands of lines of code, from kickoff to merge, using the existing test suite as the success bar.
    • Dynamic workflows are available in Claude Code for the Enterprise, Team, and Max plans.
    • Effort control arrives in claude.ai and Cowork as a setting next to the model selector that lets users choose how much effort Claude puts into a response.
    • Higher effort makes Claude think more frequently and deeply for better answers; lower effort responds faster and consumes rate limits more slowly. Effort control is available on all plans.
    • Opus 4.8 defaults to “high” effort, judged the best overall balance of quality and user experience.
    • On coding tasks, the default effort spends a similar number of tokens as Opus 4.7’s default but delivers better performance, so quality rises without a token penalty.
    • Users can select “extra” (called “xhigh” in Claude Code) or “max” to spend more tokens for stronger results, and Anthropic recommends “extra” for difficult tasks and long-running asynchronous workflows.
    • Rate limits in Claude Code were increased to accommodate the higher token usage of the higher effort levels.
    • The Messages API now accepts system entries inside the messages array, a meaningful change for agent developers.
    • That update lets developers change Claude’s instructions mid-task, adjusting permissions, token budgets, or environment context, without breaking the prompt cache or routing through a user turn.
    • Fast mode now runs at 2.5x speed and is three times cheaper than it was for previous models, priced at $10 per million input tokens and $50 per million output tokens.
    • Developers access the model as claude-opus-4-8 through the Claude API.
    • Partner Miguel Gonzalez reports Opus 4.8 scored 84% on Online-Mind2Web, a meaningful jump over both Opus 4.7 and GPT-5.5, calling it the strongest computer-use and browser-agent model his team has tested.
    • Databricks reports that, inside Genie, Opus 4.8 reasons over unstructured content like PDFs and diagrams at 61% cheaper token cost than Opus 4.7.
    • Thomson Reuters reports Opus 4.8 is the first model to break 10% overall on the all-pass standard of its Legal Agent Benchmark, the highest score recorded there.
    • Eleven partners weighed in, including Cursor, Cognition’s Devin, Databricks Genie, Thomson Reuters CoCounsel, and Hebbia, spanning coding, legal, finance, and enterprise data work.
    • Anthropic is working on models that deliver many of the same capabilities as Opus at a lower cost.
    • The company plans to release a new class of model with even higher intelligence than Opus.
    • Under Project Glasswing, a small number of organizations are already using Claude Mythos Preview for cybersecurity work, with Mythos-class models expected to reach all customers in the coming weeks once stronger cyber safeguards are in place.

    Detailed Summary

    What Claude Opus 4.8 Is

    Claude Opus 4.8 is an upgrade to Anthropic’s Opus class of models, building on Opus 4.7 with improvements across benchmarks covering coding, agentic skills, reasoning, and practical knowledge-work tasks. Anthropic describes the result as “a more effective collaborator” while characterizing the release overall as “a modest but tangible improvement on its predecessor.” The model is available today, everywhere, and developers call it as claude-opus-4-8 via the Claude API. The announcement includes a comparison table against the predecessor and other models, though the per-cell numbers in that table are published as an image and are not reproduced here as text.

    Honesty: The Headline Improvement

    Anthropic singles out honesty as one of the most prominent improvements in Opus 4.8. All of the company’s models are trained to be honest, which includes avoiding claims they cannot support. A persistent problem with AI models generally is that they sometimes jump to conclusions, confidently claiming progress despite thin evidence. Early testers report that Opus 4.8 is more likely to flag uncertainties about its own work and less likely to make unsupported claims. The most concrete measure: evaluations show Opus 4.8 is around four times less likely than its predecessor to allow flaws in code it has written to pass unremarked. For agentic and unattended use, this self-skepticism is the difference between a model that reliably tells you when something went wrong and one that quietly ships a broken result.

    Alignment Assessment

    A detailed alignment assessment was run before release. On the positive side, the Alignment team concluded that Opus 4.8 “reaches new highs on our measures of prosocial traits like supporting user autonomy and acting in the user’s best interest.” On the risk side, misaligned behavior such as deception or cooperation with misuse occurs at rates substantially lower than Opus 4.7, and similar to Anthropic’s best-aligned model, Claude Mythos Preview. The full alignment assessment and the pre-deployment safety tests are published in the Claude Opus 4.8 System Card, which also contains the complete benchmark table and wider evaluations.

    Dynamic Workflows in Claude Code

    Launching today as a research preview in Claude Code, dynamic workflows let Claude plan the work and then run hundreds of parallel subagents in a single session. With Opus 4.8, those agents can run even longer than before, and Claude verifies its outputs before reporting back rather than reporting unchecked results. The showcase example is a codebase-scale migration: Claude Code with Opus 4.8 can carry out migrations across hundreds of thousands of lines of code, all the way from kickoff to merge, using the existing test suite as its bar for success. Dynamic workflows are available in Claude Code for the Enterprise, Team, and Max plans.

    Effort Control

    Effort control arrives in claude.ai and Cowork as a setting alongside the model selector that lets users choose how much effort Claude puts into a response. Higher effort means Claude thinks more frequently and deeply for better responses; lower effort means it responds faster and uses rate limits more slowly. Opus 4.8 defaults to “high” effort, which Anthropic judged the best overall balance of quality and user experience. On coding tasks, that default spends a similar number of tokens as Opus 4.7’s default while performing better. Users who want more can choose “extra” (called “xhigh” in Claude Code) or “max” to spend more tokens for stronger results, and Anthropic recommends “extra” for difficult tasks and long-running asynchronous workflows. To support the heavier token usage at higher effort levels, rate limits in Claude Code were increased. Effort control is available on all plans.

    Messages API Update

    The Messages API now accepts system entries inside the messages array. This lets developers update Claude’s instructions mid-task without breaking the prompt cache and without routing the update through a user turn. In practice that means you can update permissions, token budgets, or environment context while an agent is running, which is exactly the kind of statefulness a long-running autonomous process needs. It is a small specification change with significant consequences for how developers build durable agents.

    Pricing and Fast Mode

    Regular usage pricing is unchanged from Opus 4.7: $5 per million input tokens and $25 per million output tokens. The notable shift is in fast mode, where the model works at 2.5x the speed and fast mode is now three times cheaper than it was for previous models, landing at $10 per million input tokens and $50 per million output tokens. The combination of unchanged regular pricing and dramatically cheaper fast mode reshapes the latency-versus-cost calculus that has long governed how teams deploy frontier models.

    Partner Results Across Coding, Legal, Finance, and Data

    Eleven partners shared results spanning the spectrum of professional work. Miguel Gonzalez reports 84% on Online-Mind2Web, a meaningful jump over both Opus 4.7 and GPT-5.5, calling it the strongest computer-use and browser-agent model his team has tested. Databricks reports that Genie reasons over unstructured content like PDFs and diagrams at 61% cheaper token cost than Opus 4.7. Thomson Reuters reports Opus 4.8 is the first model to break 10% overall on the all-pass standard of its Legal Agent Benchmark. Cursor reports gains across every effort level on CursorBench with more efficient tool calling, and Cognition reports that Devin sees cleaner tool use, fixes to the comment-verbosity and tool-calling issues seen with Opus 4.7, and improvements over Opus 4.6. Hebbia reports strong quality with better citation precision and more token efficiency on retrieval for dense financial filings. The footnotes note that Terminal-Bench 2.1 was scored on the Terminus-2 public harness (GPT-5.5’s Codex CLI harness score is 83.4%), that OSWorld-Verified methodology changed with Opus 4.7’s score updated to 82.3%, and that on Finance Agent v2 Gemini 3.5 Flash scores 57.9%.

    What Is Next: Cheaper Models, Higher Intelligence, and Mythos

    Anthropic outlined a three-part roadmap. First, the company is working on models that provide many of the same capabilities as Opus at a lower cost. Second, it plans to release a new class of model with even higher intelligence than Opus. Third, as part of Project Glasswing, a small number of organizations are currently using Claude Mythos Preview for cybersecurity work; models of this capability level require stronger cyber safeguards before general release, and Anthropic expects to bring Mythos-class models to all customers in the coming weeks.

    Notable Quotes

    “Claude Opus 4.8 has noticeably better judgment. In Claude Code, it asks the right questions, catches its own mistakes, pushes back when a plan isn’t sound, and builds up confidence around complex, multi-service explorations before making big changes. It’s a great model to build with.”

    Tom Pritchard, Staff Engineer, in Claude Code

    “On our Super-Agent benchmark, Claude Opus 4.8 is the only model to complete every case end-to-end, beating prior Opus models and GPT-5.5 at parity on cost. For agent products in translation, deep research, slide-building, and analysis, it delivers powerful reliability.”

    Kay Zhu, Co-Founder and CTO, on the Super-Agent benchmark

    “On CursorBench, Claude Opus 4.8 exceeds prior Opus models across every effort level. Tool calling is meaningfully more efficient, using fewer steps for the same intelligence, and it carries end-to-end tasks through.”

    Michael Truell, Co-Founder and CEO, on CursorBench results

    “Claude Opus 4.8 delivers the highest score recorded on our Legal Agent Benchmark, and is the first model to break 10% overall on the all-pass standard. For substantive legal work, that’s the kind of accuracy lift that translates directly into how much real attorney work our customers can hand off with confidence.”

    Niko Grupen, Head of Applied Research, on the Legal Agent Benchmark

    “Claude Opus 4.8 feels like a major quality-of-life update over Opus 4.7: faster, easier to collaborate with, and better at carrying context and style direction across a long session. Opus 4.8 is the model I kept trusting for work where voice, taste, and technical execution all have to happen side-by-side.”

    Katie Parrott, Staff Writer, on long writing sessions

    “Claude Opus 4.8 is the strongest computer-use and browser-agent model we’ve tested, scoring 84% on Online-Mind2Web, which is a meaningful jump over both Opus 4.7 and GPT-5.5. It stays reflective and on-task in the way our customers’ agent workloads need to be reliable end-to-end.”

    Miguel Gonzalez, Tech Lead, on computer-use and browser agents

    “Claude Opus 4.8 uses tools cleanly and follows instructions with the consistency our autonomous engineering workloads need to keep running unattended. It improves on Opus 4.6 and fixes the comment-verbosity and tool-calling issues we saw with Opus 4.7. This release from Anthropic translates directly into faster capability gains for engineers building on Devin.”

    Scott Wu, CEO, on building with Devin

    “On our long-running evals, Claude Opus 4.8’s analysis was consistently higher quality than prior Opus models. It finished faster and produced richer, more information dense outputs. Overall, a noticeably better signal to noise ratio. The biggest differentiator was Opus 4.8’s tendency to proactively flag issues with the inputs and outputs of an analysis, something other models routinely missed and left to the users to catch.”

    Michael Ran, Sr. Investment Associate, on long-running analysis evals

    Claude Opus 4.8 is a quieter release than its “modest but tangible” billing suggests, because the gains land where autonomous work actually lives: a model that flags its own uncertainty, runs longer and checks itself, scales effort on demand, and stays affordable while fast mode gets cheaper. The honesty improvement alone changes the trust math for anyone deploying agents. Read Anthropic’s full announcement here.

    Related Reading

  • Gavin Baker on Orbital Compute, TSMC, Frontier AI Models, Anthropic’s Vertical Take Off, and the Coming Wafer Shortage

    Gavin Baker, founder and CIO of Atreides Management, returns to Patrick O’Shaughnessy’s Invest Like the Best for his sixth appearance. He calls the current AI moment the most extraordinary moment in the history of capitalism, walks through what Anthropic’s vertical takeoff in revenue actually means, lays out why orbital compute is closer than skeptics believe, dissects the TSMC bottleneck that may be the only thing standing between today’s market and a full-on AI bubble, and rates every hyperscaler on how they have positioned for a world where frontier model providers may stop selling API access altogether.

    TLDW

    Anthropic added eleven billion dollars of ARR in a single month, which is roughly the combined business of Palantir, Snowflake, and Databricks built over a decade. That is the setup. From there Gavin Baker covers the March and April selloff, the contrarian read that a closed Strait of Hormuz was actually bullish for American manufacturing competitiveness, why Anthropic and OpenAI multiples may be misleadingly cheap on an unconstrained run rate basis, why Elon Musk’s discipline on SpaceX valuation created a superpower of permanent access to capital, the practical engineering case for orbital compute as racks in space rather than Pentagon sized space stations, why TSMC’s capacity discipline is the single most important variable in whether the AI cycle becomes a bubble, what Terafab in Texas changes, why the Pareto frontier of AI models has flipped from Google dominance to Anthropic and OpenAI dominance in nine months, the shift from all you can eat AI subscriptions to usage based pricing and what that means for revenue scaling, Richard Sutton’s bitter lesson as the largest risk to the AI trade, why frontier tokens still capture an overwhelming share of economic value, the role of continual learning as the third great open question, why most new chip startups should not try to build a better GPU, why Cerebras did something different and hard, why disaggregated inference may extend GPU useful lives to ten or fifteen years and rescue the private credit industry, why being in the token path is the new venture filter, the new prisoner’s dilemma around releasing frontier models via API, an honest rating of Google, Meta, Amazon, and Microsoft, why personal safety is becoming a real AI era risk, and why he remains an AI optimist maximalist who believes this could be the next Pax Americana.

    Key Takeaways

    • Anthropic added eleven billion dollars of ARR in one month, more than the combined businesses of Palantir, Snowflake, and Databricks built across a decade. There is no precedent for this in the history of capitalism.
    • The SaaS and cloud revolution created between five and ten trillion dollars of value over twenty years. AI is replaying that compression on a timeline measured in months.
    • The March selloff was a drawdown driven by disagreement with price action, not invalidated thesis. That is the kind of drawdown an investor can lean into.
    • Deep Seek Monday in January 2025 was a similar setup. By the day of the selloff, AWS Asia GPU prices had already doubled, GPU availability had fallen, and it was obvious reasoning models would be vastly more compute hungry at inference. The market priced the opposite.
    • The Strait of Hormuz closing was actually positive for America. US natural gas (the primary input into US electricity, which feeds AI) fell twenty percent on Bloomberg while Asian and European natural gas doubled or tripled. American manufacturing competitiveness improved overnight.
    • The US is now the world’s largest producer and exporter of oil and gas. The economy is dramatically less energy intensive than in the 1970s. The shortage trauma comparison does not hold.
    • Tech as a sector traded as cheaply versus the rest of the market in early April as at any point in the last ten years, into the single most bullish moment for AI fundamentals on record.
    • Anthropic is dramatically more capital efficient than OpenAI, having burned roughly eighty percent less to reach a similar revenue scale. They have very different structural returns on invested capital.
    • Anthropic at roughly nine hundred billion for fifty billion of ARR (growing a thousand percent) is striking. Adjusted for compute constraint, the unconstrained run rate could be one hundred fifty to two hundred billion, putting the implied multiple closer to five times.
    • Claude Opus generates roughly seventy percent fewer tokens for the same question than previously, with token quantity tied to answer quality. Subscribers on flat-fee plans are getting a lobotomized model.
    • Elon Musk’s superpower is twenty years of making investors money. He never pushes valuation. SpaceX compounded low thirty percent per year for a decade because Musk treats fair pricing as a sacred covenant.
    • Capitalism will solve the watts shortage. The current bottleneck has shifted from chips and energy to zoning and political approval. Many capex decisions are paused until after the US midterms.
    • The watts shortage probably begins to alleviate in 2027 and 2028. Orbital compute solves it longer term.
    • Orbital compute is not Pentagon sized data centers in space. It is racks in space. A Blackwell rack is three thousand pounds, eight feet tall, four feet deep, three feet wide. SpaceX has shown a satellite roughly that size.
    • The satellites operate in sun synchronous orbit so solar wings (around five hundred feet per side) always face the sun and the radiator on the dark side always points to deep space.
    • Starlink V3 satellites already run at around twenty kilowatts. A Blackwell rack runs at one hundred kilowatts. SpaceX engineers express genuine confidence they have already solved cooling and radiator design at these scales.
    • Racks in space are connected with lasers traveling through vacuum, the same lasers already on every Starlink. SpaceX operates the world’s largest satellite fleet and, via xAI Colossus, the world’s largest data center on Earth.
    • Inference will move to orbit. Training will stay on Earth for a long time. Terrestrial data centers remain valuable for the rest of an investor’s career.
    • The wafer bottleneck is structural and political. TSMC is essentially Taiwan’s GDP, water, and electricity. The leaders see themselves as inheritors of Morris Chang’s sacred legacy and they do not behave like a Western public company.
    • Jensen Huang has never had a contract with TSMC. The relationship is run on handshakes and the assumption that things will be fair over time.
    • If TSMC did everything Jensen wanted, Nvidia could be selling two to three trillion dollars of GPUs in 2026 and 2027. TSMC’s discipline is the single largest factor preventing a true AI bubble.
    • Historically, foundational technologies always get a bubble. Railroads, canals, the internet. The current AI buildout is overwhelmingly funded out of operating cash flow, GPUs are running at one hundred percent utilization, and that is fundamentally different from the year 2000 fiber overbuild.
    • If one of Intel or Samsung Foundry catches up at the leading node, the other will follow, and TSMC’s discipline collapses. Watch TSMC capacity decisions to predict a bubble.
    • Terafab, the SpaceX and Tesla joint venture to build the world’s largest fab in America, has a partnership with Intel that grants access to fifty years of institutional foundry knowledge. The A teams at ASML, KLA, Lam Research, and Applied Materials will follow Elon’s reputation in hardware engineering.
    • The hiring playbook for Terafab includes building Taiwan Town, Japan Town, and Korea Town next to the fab. Recruit the engineers and import their families, their restaurants, and their staff.
    • Frontier tokens still capture an overwhelming share of all economic value created at the model layer. This is surprising and is one of the three big open questions for AI investing.
    • The Pareto frontier of intelligence versus cost has flipped. Nine months ago Google’s TPU dominated every point on the frontier. Today Anthropic and OpenAI dominate, with Grok 4.3 on the frontier and Gemini 3.1 hanging on.
    • Google’s conservative TPU V8 design (partly an attempt to reduce dependence on Broadcom and Nvidia) is the leading explanation for the loss of per token cost leadership.
    • AI pricing is shifting from all you can eat to usage based, mirroring the cellular and long distance industries. Cellular stopped being a great growth industry when it went all you can eat. AI just made the opposite move.
    • OpenAI and Anthropic together could exceed two hundred billion in ARR this year if compute keeps coming online and frontier token pricing holds.
    • The two hundred fifty dollar a month consumer AI plan is no longer enough to evaluate frontier capability. Enterprise plans with usage based billing are required because rate limits are now severe.
    • The three biggest open questions for AI investors are: violation of the bitter lesson via ASI or human ingenuity, whether frontier tokens keep commanding their premium, and when continual learning arrives.
    • Today’s continual learning is crude reinforcement learning during mid training on verifiable tasks. True continual learning means weights updating dynamically, like a human who learns the first time they touch fire.
    • Trying to build a better GPU is a losing strategy. Jensen will copy any one to three percent share design. Startups should target one percent share, do something different, and make it hard enough that Nvidia cannot fast follow.
    • Disaggregated inference (separating prefill and decode) opens new design canvases. Prefill is memory capacity bound. Decode is memory bandwidth bound. Each can be optimized independently.
    • Cerebras did something different and hard with wafer scale computing. Three generations of chips and real grit to get there.
    • Disaggregation of inference may stretch GPU useful lives to ten or fifteen years, dropping financing costs from low sevens to five or six percent, mathematically lowering the cost of the AI buildout and likely saving the private credit industry from its SaaS loan exposure.
    • Sellers of shortage outperform buyers of shortage. But owning the largest installed base of what is currently in shortage (hyperscaler CPU fleets, for example) is also a strong position.
    • Most of the economic value at the application layer of AI has been destroyed, not created. The exceptions are companies in the token path or in niches small enough that frontier labs ignore them.
    • Coding may be the shortest path to ASI. If you can write code, you can write code that does anything. Cursor, Cognition, and Anthropic correctly focused on it.
    • Jensen could probably get close to the frontier with his own Nemotron family of models whenever he wants. The fact that he chooses not to is a strategic decision about not commoditizing his customers.
    • The new prisoner’s dilemma in AI is whether frontier labs release their best model via API. If everyone agrees not to, Chinese open source falls behind. If anyone defects, the defector pulls ahead on revenue and resources, forcing everyone else to defect.
    • Google still owns the largest compute installed base. Without TPU’s prior cost advantage, this matters more. YouTube data has real value in a world of robotics. GCP is going crazy.
    • Meta deserves credit for becoming AI first internally faster than any other internet giant. Musa, their first MSL model, is impressively close to the Pareto frontier.
    • Amazon is strong because of Trainium and robotics driven retail P&L efficiency. Nova is better than it gets credit for.
    • Microsoft flinched on capex in early 2025 and lost position. Satya Nadella’s current decision to use Microsoft compute for Microsoft products rather than reselling to OpenAI is a courageous and probably correct call, even at the cost of an eight hundred dollar stock price.
    • The hyperscalers most engaged with startups are Amazon and Nvidia by a mile, followed by Google. Broadcom is the favorite ASIC partner. AMD, Microsoft, and Meta have minimal startup engagement and that will cost them as the best teams are now at startups.
    • Personal safety in an AI era requires a family or company safe word that cannot be socially engineered. Deepfake voice and video extortion at the speed of FaceTime is already feasible.
    • Ukraine is winning largely on the back of having the best battlefield AI outside America and Israel. Adversaries are starting to internalize what AI dominance means geopolitically.
    • An optimistic read is that this becomes a new Pax Americana, the way the post 1945 American nuclear monopoly was used to rebuild Germany and Japan rather than dominate.
    • AI cured a friend’s daughter’s rare disease by spinning up a research effort that identified a market drug capable of impacting her condition. That is the upside that keeps Gavin an AI optimist maximalist.

    Detailed Summary

    The most extraordinary moment in the history of capitalism

    Gavin’s framing of the current moment is unusually direct. Anthropic added eleven billion dollars of annual recurring revenue in a single month. The three highest profile SaaS companies of the last decade plus, Palantir, Snowflake, and Databricks, took a decade and tens of thousands of employees collectively to build the combined business that Anthropic added in thirty days. He has been investing through every major tech cycle and says there is no historical analog. Not the dotcom era, not the cloud transition, not mobile. This is its own thing.

    The market response, then, was peculiar. The NASDAQ sold off into the single most bullish moment for AI fundamentals on record. Tech traded at roughly its widest discount versus the rest of the market in a decade. Investors who said they wished they had bought into AI during 2022, during COVID, or during Deep Seek Monday got the same valuation setup again in early April, this time with an even clearer inflection.

    Why the Strait of Hormuz closing was secretly bullish for America

    One reason the macro fear in March may have been mispriced is that the same geopolitical event that drove the selloff was, in practice, a relative benefit to the United States. American natural gas, the input into American electricity, which is the input into American AI training and inference, fell roughly twenty percent. Asian and European natural gas prices doubled or tripled. The US emerged with sharply improved relative manufacturing competitiveness, which is exactly what the current administration cares about.

    The 1970s comparison does not hold. The US economy is dramatically less energy intensive, it is now the world’s largest producer and largest exporter of oil and gas, and there are no shortages, only price moves. That backdrop made it easier for disciplined investors to stay focused on AI fundamentals through the volatility.

    Anthropic and OpenAI valuations on an unconstrained run rate

    Anthropic at roughly nine hundred billion for fifty billion of ARR sounds rich until you adjust for the fact that the company is severely compute constrained. Gavin estimates that, unconstrained, Anthropic might be at one hundred fifty to two hundred billion in run rate revenue, putting the implied multiple closer to five times. He also points out that Claude Opus now generates roughly seventy percent fewer tokens for the same question than it used to. Token quantity correlates with answer quality, and Anthropic is rate limiting and shrinking outputs to ration capacity across its user base.

    Anthropic and OpenAI are also structurally very different. Anthropic has burned around eighty percent less cash than OpenAI to reach a comparable revenue scale. That implies very different long term returns on invested capital, though OpenAI has done a better job locking in compute and Sarah Friar is one of the most exceptional CFOs Gavin has worked with.

    Why neither lab is raising at a three trillion dollar valuation

    The answer Gavin gives is that both labs are deliberately leaving valuation on the table the way Elon has done for two decades. SpaceX compounded at low thirty percent annually for a decade because Elon never pushed price. The result is a permanent superpower of access to capital. Investors trust him because they have made money with him for twenty years. That is a moat that compounds with every round.

    Anthropic could probably raise at a one hundred percent premium to its rumored latest mark. They are choosing not to. In an uncertain world (Ukraine, Russia, Iran, Taiwan), preserving the ability to raise more capital later at fair prices is more valuable than maximizing this round.

    Watts and wafers, the two real constraints

    Capitalism is solving the watts problem. The leading PE infrastructure investors now say zoning and political approval, not chips or energy, are the gating factors. Companies are deferring big capex announcements until after the US midterms. Turbine capacity is being doubled at the manufacturers. Companies like Boom Aerospace are repurposing jet engines for grid use. Watts probably ease meaningfully in 2027 and 2028 and then orbital compute does the rest.

    Wafers are the harder problem because they live in Taiwan, run on handshakes, and depend on a corporate culture that does not respond to public market incentives. TSMC is essentially the GDP, water consumption, and electricity consumption of Taiwan. Its leadership treats the company as the legacy of Morris Chang. The Silicon Shield doctrine is real and internal.

    Orbital compute as racks in space

    The biggest mental update Gavin asks listeners to make is to stop picturing data centers in space as Pentagon sized space stations. A Blackwell rack is three thousand pounds and roughly the size of a refrigerator. SpaceX has shown a concept satellite of about that size. Solar wings extend five hundred feet to each side and the radiator extends hundreds of feet behind, both possible because the orbit is sun synchronous and the orientation is fixed relative to the sun.

    SpaceX engineers Gavin has spoken to at Starbase express genuine confidence that they have solved cooling at these power levels. They have. Starlink V3 satellites already operate at twenty kilowatts. A Blackwell rack is one hundred kilowatts. The same company operates the world’s largest satellite fleet and the world’s largest data center on Earth via xAI Colossus. The racks are connected to each other with lasers traveling through vacuum, technology already deployed in every Starlink. The naysayers, Gavin observes, are armchair skeptics and Larry Ellison’s response (he is out there landing rockets, no one else is) is the right frame.

    Terafab in Texas and the threat to TSMC’s discipline

    Terafab, the SpaceX and Tesla joint venture, intends to be the largest fab in the world. The partnership with Intel grants access to fifty years of foundry institutional knowledge, allowing Terafab to start three to five quarters behind the leading node rather than fifteen years behind. The A teams at the semicap equipment companies (ASML, KLA, Lam Research, Applied Materials) will follow Elon’s reputation in hardware engineering the same way they followed TSMC twenty years ago when Intel stumbled.

    The talent strategy is the part most observers underestimate. Recruit the best engineers globally, then import their families, their restaurants, their staff. Build Taiwan Town, Japan Town, and Korea Town next to the fab. Optimize the human experience for the people whose work matters. Intel and Samsung do not think that way.

    Bubble watch and the year 2000 comparison

    Every foundational technology in modern history has had a bubble. Railroads, canals, the internet. Carlota Perez documented why. Markets correctly identify the importance, diversity of opinion collapses, supply gets ahead of demand, the bubble crashes. The current cycle has two important differences. The buildout is overwhelmingly funded out of operating cash flow, not debt. Every GPU is running at one hundred percent utilization, while at the peak of the fiber bubble ninety nine percent of fiber was unused.

    TSMC discipline is the single largest reason a bubble has not formed. If Jensen could buy everything TSMC could theoretically make, Nvidia could sell two to three trillion dollars of GPUs in 2026 and 2027. At some point that becomes more than the market can absorb. If Intel or Samsung Foundry catches up at the leading node, the other will too. TSMC’s pricing discipline collapses and the bubble starts.

    The Pareto frontier and the loss of Google’s cost advantage

    The most important chart in AI is the Pareto frontier of model intelligence versus per token cost. Nine months ago, Google’s TPU based models dominated every point on it. OpenAI, Anthropic, and xAI sat inside the frontier. Today the frontier is dominated by Anthropic and OpenAI, with Grok 4.3 on the frontier and Gemini 3.1 hanging on by subsidization more than economics. The most likely cause is Google’s conservative TPU V8 design, an attempt to reduce dependence on Broadcom and Nvidia that sacrificed per token economics.

    The bitter lesson, frontier tokens, and continual learning

    Three open questions dominate AI investing. The first is whether Richard Sutton’s bitter lesson (more compute beats human algorithmic cleverness) gets violated by ASI itself optimizing for efficiency. Closer observers of AI are more skeptical of a violation. Gavin thinks ASI’s first move will be to make itself more efficient and more resourced, which is technically a temporary violation.

    The second is whether frontier tokens keep capturing the overwhelming share of economic value at the model layer. Today they do, surprisingly. Gemini 3.1 Pro was mindblowing nine months ago and is intolerable today. The third is when continual learning arrives. Today’s models need a million fire touches to learn what a human learns from one. True continual learning would mean dynamic weight updates in real time and would produce a fast takeoff.

    From all you can eat to usage based AI pricing

    AI is shifting from flat fee plans to usage based pricing. The historical analogy is cellular and long distance. Both stopped being great growth industries when they went all you can eat. AI just made the opposite move. The consequence is that flat fee subscribers, even on premium consumer plans, get a rate limited and token throttled version of the frontier model. Enterprise plans with usage based billing are now required to evaluate true capability. Gavin thinks the combination of new compute coming online and usage based pricing is what gets OpenAI and Anthropic past two hundred billion in combined ARR this year.

    Chip startups, prefill decode disaggregation, and Cerebras

    Trying to build a better GPU is the wrong move. The four scaled players (Nvidia, AMD, Trainium, TPU) have copy capability for any one to three percent share design that looks attractive. The good news for startups is that disaggregated inference (separating prefill and decode) opens a richer design canvas. Prefill is memory capacity bound. Decode is memory bandwidth bound. Each can be optimized independently. Andrew Fox’s analogy is a British naval ship of the eighteenth century. Prefill is loading the cannon. Decode is firing it.

    Cerebras is the model. Wafer scale computing is genuinely different and genuinely hard. It took three generations of chips to get right. Andrew Feldman and his team had the grit to keep going through chip one being a failure. The design has a high ratio of on chip compute and memory relative to shoreline IO, which is why Cerebras is now experimenting with putting an optical wafer on top of the compute wafer to solve scale out.

    GPU useful lives and the rescue of private credit

    One of the strongest claims in the conversation is that disaggregated inference will stretch GPU useful lives to ten or fifteen years. The skeptical narrative (GPUs are obsolete in two years, companies are cooking their depreciation books) is wrong. You can put a Cerebras system or Groq LPU in front of older Hopper or Ampere parts, use them only for prefill, and run them until they physically melt. Private credit, which is in pain from SaaS loans and which underwrote GPU loans on three to four year lives, may be saved by this.

    If GPU financing rates can come down from low sevens to five or six percent, the mathematics of the AI buildout improves materially. That is a structural tailwind that compounds for years.

    The application layer, the token path, and a new prisoner’s dilemma

    Trillions of dollars of value have been destroyed at the application layer, not created. Cursor and Cognition are the rare scaled exceptions, and they got there by focusing on coding very early. As Amjad Masad noted, coding is plausibly the shortest path to ASI because a coding agent can write itself into any new domain. Jamin Ball’s frame is that the new venture filter is whether the company is in the token path. Data Bricks is. Most application layer startups are not.

    Jensen could probably get close to the frontier with Nemotron whenever he wants, and the strategic question of whether to do that is a new prisoner’s dilemma. If every frontier lab agrees not to release best models via API, Chinese open source falls steadily behind. If anyone defects, the defector gains revenue and resources, and everyone else has to defect. The same dynamic exists between TSMC, Intel, and Samsung. If Nvidia or AMD ever truly used an alternative foundry, that foundry would catch up rapidly.

    Rating the hyperscalers

    Google has the largest compute installed base, the YouTube data that matters in a robotics world, and a search business that prints. Their loss of TPU cost leadership is the surprise of the year. If Google IO in five days does not produce a leapfrog model, the Nvidia centric narrative gets even stronger.

    Meta deserves real credit. Zuckerberg made Meta AI first internally faster than any other internet giant, paid up for the talent contracts when no one else would, and shipped Musa as a first model from MSL that is close to the Pareto frontier. Amazon is well positioned on Trainium, robotics in retail, and a Nova model line that is better than it gets credit for. Microsoft flinched on capex in early 2025 and lost position. Satya Nadella’s current decision to use Microsoft compute for Copilot rather than reselling to OpenAI is courageous and probably correct, even at the cost of stock price.

    The most interesting cross hyperscaler metric is startup engagement. Nvidia and Amazon engage deeply with startups. Google is next. Broadcom is the favored ASIC partner. AMD, Microsoft, and Meta have minimal startup engagement, which Gavin believes will cost them as the best teams now sit at startups.

    Personal safety, geopolitics, and the Pax Americana case

    The closing section turns darker. Personal safety in an AI era requires a family or company safe word that cannot be socially engineered. Deepfake voice and video extortion via something that looks exactly like your child calling on FaceTime is already feasible. Political violence against AI leaders is a real concern. Geopolitically, Ukraine is winning largely because it has the best battlefield AI outside America and Israel. How adversaries respond to that asymmetry is the next great variable.

    Gavin’s optimistic frame is the Pax Americana. After 1945 the US had a nuclear monopoly and could have controlled the world. Instead it rebuilt Germany and Japan, both of which became the most reliable American allies for the next eighty years. If AI dominance plays out similarly, this is a generationally positive story rather than a destabilizing one. The personal anecdote that closes the conversation is a friend whose daughter was diagnosed with a rare genetic condition. He spun up agents, identified a drug already on the market that addresses her mutation, and her life is immeasurably different because of AI. That is the upside.

    Thoughts

    The Anthropic eleven billion in a month framing is the kind of stat that resets priors. The right way to interpret it is not as a one off but as a measure of how fast value can compound when the underlying technology improves on a curve steeper than the ability of the rest of the economy to absorb it. The skeptical question is whether that ARR is durable or whether it is heavily tied to a customer base of other AI companies that are themselves on a single venture funded year of runway. The bullish answer is that frontier coding, frontier research, and frontier enterprise tasks are not going to stop being valuable, and Anthropic is the best at all three. Both can be true. The number is still extraordinary.

    The argument that TSMC discipline is the only thing preventing a bubble is the analytically tightest part of the conversation. The implied trade is to watch TSMC capacity additions like a hawk and to be more, not less, cautious if Intel Foundry or Samsung Foundry ever announce real share at the leading node. The Terafab thesis is more speculative but more interesting. If Elon’s talent recruiting playbook works and the Intel partnership gives Terafab a real seat at the table within five years, the geometry of the global semiconductor industry shifts in a way that is bullish for American manufacturing, bullish for power and water infrastructure in Texas, and ambiguous for TSMC itself.

    The Pareto frontier discussion deserves more attention than it usually gets. Pricing leadership in AI is not a vanity metric. It determines who can subsidize free tier usage, who can absorb compute shortages, who can ship cheaper enterprise plans, and ultimately whose model becomes the default for any given workload. Google losing per token leadership in nine months is one of the most under analyzed events in the sector and it explains a lot about why Anthropic and OpenAI are growing the way they are. If Google IO does not produce a leapfrog model, the implied verdict on TPU V8 design choices gets a lot harsher.

    The application layer destruction point is worth sitting with. Founders building on top of frontier models are competing in a world where the model itself moves faster than any moat they can build, where the model lab can absorb their niche if it gets interesting, and where the only protection is either deep token path integration or a niche so small the lab does not bother. That is a much harsher venture environment than the early SaaS era. The compensating opportunity is that one human can now run a hundred agents, so the ceiling on what a small team can build is correspondingly higher. The bet is that productivity per founder rises faster than competitive pressure from the labs. We will find out.

    The orbital compute pitch is the section that will polarize listeners. The naive read is that this is science fiction. The closer read is that every component (sun synchronous orbit, laser interconnect, twenty kilowatt satellite buses, ten thousand satellite manufacturing cadence, full rocket reusability) already exists. The remaining engineering problems are repair, maintenance, and radiator scale, all of which are real but tractable on a five to ten year horizon. The strategic implication is that the political and zoning ceiling on terrestrial data centers becomes less binding if orbital compute is a credible alternative for inference workloads. The investor implication is that being short the watts and cooling complex on a five year horizon is a real trade, not a meme.

    Watch the full conversation here.