As of November 2025, X has rolled out a significant transparency update. The platform now displays every account’s detected base country directly on their profile. This new feature, nestled under the “About this account” section, is designed to verify the authenticity of users and combat misinformation.
Whether you want to verify a source’s location or adjust your own privacy settings, here is everything you need to know about the new location display.
How to View Someone’s Location on X
You do not need to follow a user to see their location data. It is visible to everyone by default. Here is how to access it in just three taps:
Open the user’s profile page.
Tap the “Joined” date located at the top of their profile (next to the calendar icon).
Look for the line reading “Account based in [Country].”
What the Location Status Means
X uses IP activity to detect location, but users also have manual control over how specific this information is. Here is a breakdown of what you might see on a profile:
Displayed Text
What It Means
United States / India / etc.
This is the default setting. X has detected the precise country based on recent IP activity.
North America / Europe / etc.
The user has manually changed their settings to show a broader “Region” rather than a specific country.
“Country or region may not be accurate”
X has detected active VPN or proxy usage. This warning is automatically appended to the profile.
Location hidden
The user has opted out of the feature entirely. Note: Profiles hiding this info may be flagged or highlighted by the platform.
How to Change or Hide Your Own Location
If you are concerned about privacy, you can alter how—or if—your location is displayed. The feature is currently rolling out globally and is expected to be fully enforced within weeks.
To change your settings, follow this path:
Settings & Privacy → Privacy and safety → About your account → Account location display
From there, you can select one of three options:
Country: The default, precise view.
Region: A broader, less specific view.
Hide entirely: Removes the data from your profile.
Live from The Venetian: The Besties break down the Epstein file release, the massive margins of Tether, the Michael Burry vs. Nvidia debate, and a masterclass in risk with Alan Keating.
In this special live episode recorded during the F1 weekend in Las Vegas, the “Besties” (Chamath Palihapitiya, Jason Calacanis, David Sacks, and David Friedberg) reunite in person. The agenda is packed: political intrigue surrounding Jeffrey Epstein, the financial dominance of stablecoins, technical debates on AI chip accounting, and high-stakes poker strategy.
TL;DR: Executive Summary
The US government has voted nearly unanimously to release the Epstein files, leading the hosts to speculate that the lack of leaks points to intelligence agency involvement rather than political dirt on Donald Trump. Chamath details a meeting with Tether CEO Paolo Ardoino, revealing a business holding over $100 billion in US Treasuries with profit margins potentially exceeding 95%. The group then debates Michael Burry’s short position on Nvidia, with Friedberg defending the “useful life” of AI chips under GAAP accounting. Finally, poker legend Alan Keating joins to discuss “soul reading” opponents and mastering fear in high-stakes games.
Key Takeaways
The Epstein Intelligence Theory: The hosts argue that if the files contained damaging information on Donald Trump, it would have been leaked during the Biden administration. The prevailing theory discussed is that Epstein may have been an intelligence asset (CIA/Mossad/Russia), explaining the long-standing secrecy.
Tether is a Financial Juggernaut: Tether holds approximately $135 billion in US Treasuries and operates with roughly 100 employees. Chamath estimates the business runs at 95%+ margins, effectively exporting US dollar stability to developing nations while capturing massive interest yields.
Nvidia vs. Michael Burry: “The Big Short” investor Michael Burry is shorting the sector, arguing tech companies are “cooking the books” by depreciating AI chips over 6 years when they become obsolete in 3. Friedberg counters that chips retain a “useful life” for inference and background tasks long after they are no longer top-of-the-line.
Google Gemini 3: Google has regained the lead on LLM benchmarks with Gemini 3. The conversation highlights a shift toward proprietary silicon (TPUs) and a fragmented chip market, posing a potential long-term risk to Nvidia’s dominance.
The “Oppenheimer” Moment: David Friedberg reveals he decided to return as CEO of Oho after watching the movie Oppenheimer, realizing he needed to be an active operator rather than a passive board member.
Detailed Episode Breakdown
1. The Epstein Files Release
In a stunning bipartisan move, the House and Senate voted nearly unanimously to release the Epstein files. The Besties analyzed why this is happening now. Sacks and Chamath suggested that because Epstein was the “most investigated human on earth,” any compromising information regarding Trump would likely have been weaponized politically by now.
The discussion pivoted to the source of Epstein’s wealth. Chamath noted Epstein managed money for billionaires and charged inexplicable fees for “tax advice”—such as a documented $168 million payment from Apollo’s Leon Black. The hosts speculated that Epstein likely functioned as a spy or asset for intelligence agencies, which would explain the protective layer surrounding the files for so long.
2. Tether and the Stablecoin Boom
Chamath shared insights from a dinner with Tether CEO Paolo Ardoino. Tether’s financials are staggering: approximately $135 billion in US Treasuries and billions more in Bitcoin and gold.
The hosts discussed the utility of stablecoins in high-inflation economies, where locals use USDT to preserve purchasing power. Because Tether earns the interest on the backing treasuries (rather than passing it to the coin holder), and operates with a lean team, the company generates billions in pure profit. Sacks noted that future US regulations might eventually force stablecoin issuers to share that yield with users, but for now, it remains one of the most profitable business models in the world.
3. Accounting Corner: Is Nvidia Overvalued?
Michael Burry is shorting the semiconductor sector, claiming companies are inflating earnings by depreciating Nvidia chips over 6 years despite rapid technological obsolescence.
Friedberg launched a segment dubbed “Accounting Corner” to rebut this. He explained that under GAAP standards, an asset’s useful life is determined by its ability to generate revenue, not just its technological superiority. Even if an H100 chip isn’t the fastest on the market in year 4, it can still run inference models or handle lower-priority compute tasks, justifying the longer depreciation schedule. Chamath added that tech giants monitor “output tokens” closely; if a chip wasn’t profitable, they would simply turn it off.
4. Poker Strategy with Alan Keating
The episode concluded with Alan Keating, a high-stakes poker player famous for his loose, aggressive style. Keating explained his philosophy, which relies less on “solvers” (GTO strategy) and more on “soul reading”—navigating the fear and psychology of the table.
He broke down a famous hand where he beat Doug Polk with a 4-2 offsuit, explaining that he sensed fear in Polk’s betting patterns on the turn. Keating described his approach as finding “beauty in the chaos” and dragging opponents into “deep water” where they are uncomfortable and prone to errors.
Editorial Thoughts
This episode marked a distinct shift in the podcast’s tone regarding crypto, moving from general skepticism to a recognition of the sheer scale and utility of stablecoins like Tether. The “Accounting Corner” segment, while technical, provided critical context for investors trying to value the AI stack—suggesting the AI boom has more fundamental accounting support than bears like Burry believe. Finally, the live format from Las Vegas brought a looser, more energetic dynamic to the conversation, highlighting the chemistry that makes the show work.
In a recent discussion on the Uncapped podcast with Jack Altman, Robinhood co-founder and CEO Vlad Tenev opened up about the company’s transition from a trading platform to a “financial super app.” Tenev discussed the explosion of prediction markets, the role of AI in creating “vibe trading,” and his vision for tokenizing private assets to help retail investors capture value earlier.
TL;DR
Robinhood is aggressively expanding beyond simple stock trading. Vlad Tenev highlights three major frontiers: the rise of prediction markets as “truth machines,” the use of AI to create autonomous “vibe trading” experiences, and the tokenization of private assets to allow everyday investors access to companies like SpaceX or OpenAI before they go public.
Key Takeaways
From App to Ecosystem: Robinhood no longer views itself merely as a trading platform but as a “financial home” and super app, encompassing banking, credit cards, and retirement accounts.
Prediction Markets are Booming: Tenev views prediction markets not just as speculation, but as “truth machines” that offer cleaner data than traditional polling or media. Robinhood’s volume in this sector has seen massive growth.
“Vibe Trading”: Tenev coined the term “vibe trading” to describe a future where AI agents manage a user’s portfolio based on high-level intent, risk tolerance, and personal goals rather than manual trade execution.
Solving the Private Equity Gap: Tenev argues that the biggest inequity in modern markets is that value now accrues in private markets (e.g., SpaceX, OpenAI) rather than public ones. He believes tokenization is the solution to give retail investors access.
Generational Shifts: Contrary to stereotypes, Gen Z is opening retirement accounts as early as 19 years old, signaling a shift toward financial conservatism compared to millennials.
Detailed Summary
The Evolution of the Brokerage
Tenev traces the history of the online brokerage from the deregulation of commissions in 1975 (the “Mayday” event that birthed Charles Schwab) to the mobile-first revolution led by Robinhood. While early digital brokers like E-Trade catered to Gen X, Robinhood capitalized on two shifts: the ubiquity of mobile phones and the infrastructure changes brought by high-frequency trading, which lowered costs enough to offer commission-free trading.
Today, Robinhood generates over a billion dollars in revenue across multiple business lines, aiming to be the primary financial institution for its users.
Prediction Markets: The “Truth Machines”
One of the fastest-growing segments for the company is prediction markets. Tenev notes that the 2024 Presidential Election was a “Big Bang” moment for the industry, validating these markets as superior forecasting tools compared to traditional polls.
He argues that because participants have “skin in the game,” prediction markets filter out noise and bias, acting as “truth machines.” Beyond politics, this is expanding into sports and entertainment, which Tenev views as an inevitability in an economy where AI automates traditional labor.
Tokenization and Private Markets
Tenev expressed deep concern regarding where economic value is created today versus thirty years ago. When Microsoft and Apple went public, they were valued in the low billions, allowing public market investors to capture the majority of their growth. Today, companies like SpaceX or OpenAI may reach trillion-dollar valuations while still private, shutting out retail investors.
His solution is tokenization. Similar to how stablecoins operate, Tenev envisions a structure where private securities are held in a “bucket” while tokens representing them trade freely 24/7 on a blockchain. This would democratize access to private equity, a move he sees as the eventual end-state of capital markets.
AI and the Era of “Vibe Trading”
Robinhood is heavily integrating AI into its operations, achieving high deflection rates in customer support and increased coding output from engineering. However, the consumer-facing future is what Tenev calls “Vibe Trading.”
In this model, the user interface shifts from manual execution to intent-based directives. A user might tell an AI agent their risk appetite, long-term goals, and interests, and the agent—acting as a “financial home”—executes the strategy. Tenev believes this will also solve mundane friction points, such as AI agents automatically handling the paperwork to switch bank accounts.
Thoughts on the Interview
Vlad Tenev’s commentary suggests a significant pivot in Robinhood’s brand identity. Originally seen as the disruptor that “gamified” trading, the company is now positioning itself as the mature “financial super app” for a generation that is aging into wealth.
The most compelling insight is the focus on tokenization. Tenev correctly identifies that the “public market” is no longer the primary engine of wealth creation for early-stage innovative companies. If Robinhood can successfully navigate the regulatory hurdles to tokenize private equity (essentially breaking down the walls of the accredited investor requirements via technology), they wouldn’t just be a brokerage; they would fundamentally alter the structure of modern capitalism.
Furthermore, the concept of “Vibe Trading” aligns with the broader tech trend of “agentic AI.” It moves the user value proposition from “we give you the tools to do it yourself” to “we have the intelligence to do it for you,” which may appeal to a broader demographic than active traders.
November 20, 2025 – NVIDIA just delivered the most dominant quarter in the history of tech and told the world the next one will be even bigger. The market is partying like it’s 2021.
TL;DR
Revenue $57.01B (+62% YoY, beat by ~$1.8–2B)
Data Center $51.2B (+66% YoY, +$10B sequentially) – now 90% of total revenue
GAAP EPS $1.30 (+67% YoY)
Q4 guidance $65B (±2%) – obliterates street $61.98B (some buyside whispers were $75B → Jensen sandbagging again)
Blackwell sales “off the charts”, cloud GPUs completely sold out for the foreseeable future
CFO Colette Kress confirmed ≈$500B Blackwell + Rubin revenue visibility 2025–2026 (analysts now calling it $500B pipeline through FY2027)
Gross margin 73.6% (tiny miss due to Blackwell ramp costs), guided back to 75.0% next quarter
Free cash flow $22.1B in a single quarter
Top 4 customers = 61% of revenue (22% / 15% / 13% / 11%) – concentration risk is real but demand makes it a feature
Stock ripped +5.5% after-hours → +$220B+ market cap in minutes, lifting entire AI complex
Key Takeaways
Demand is not slowing — it’s compounding. Jensen: “Compute demand keeps accelerating and compounding across training and inference — each growing exponentially. We’ve entered the virtuous cycle of AI.”
Blackwell ramp is unprecedented – already the majority of new Data Center mix, sold out for months, driving the entire $10B sequential jump
Gaming ($4.3B) and Automotive ($592M) missed estimates → literally nobody cares when Data Center grew $10B in one quarter
Customer concentration: Four hyperscalers = 61% of revenue. Everyone knows who they are. Everyone also knows they can’t build without NVIDIA
Margins dipped to 73.6% only because of Blackwell complexity/HBM costs – guided 75% next quarter, street relieved
Balance sheet is absurd: $60.6B cash + $22.1B quarterly FCF. Berkshire is only ~$320B ahead
Physical AI multi-trillion opportunity already “multi-billion” today
Detailed Summary
NVIDIA printed $57.01 billion in a single quarter — a number larger than the entire annual revenue of 99% of public companies. Data Center alone did $51.2 billion (+66% YoY, +$10 billion sequentially). Let that sink in.
Blackwell is not “ramping” — it’s exploding. It is already the majority of new Data Center revenue and cloud providers are in a literal bidding war for every wafer. Jensen was blunt: “Blackwell sales are off the charts, and cloud GPUs are sold out.”
Yes, Gaming and Automotive missed estimates (who cares), Pro Visualization crushed it (+56% YoY), but the only number that matters is the $500 billion in confirmed Blackwell + Rubin orders the company can already see through calendar 2026 (Bloomberg Intelligence now calling it $500B pipeline through fiscal 2027).
China export restrictions? Effectively $0 impact in guidance. The rest of the planet is making up for it and then some — sovereign AI factories, enterprises, everyone is building.
Networking (Spectrum-X + InfiniBand) up ~162% YoY to $8.2B+ — the hidden monster line item nobody talks about.
Market & Analyst Reaction
Initial spike was +4%, then kept climbing → closed extended trading up ~5.5%, adding north of $220 billion in market cap. Entire AI food chain ripping: CoreWeave +4%, Nebius +4%, AMD +2%, Micron +2%, Broadcom +2%, Super Micro +8%.
Goldman Sachs (James Schneider) first note post-earnings:
“Strong quarter with upside to guidance should provide relief for the stock… We expect the stock to trade higher following a stronger quarter and guidance relative to the Street.”
X was pure euphoria last night – here are some of the top posts (all >5K likes):
https://x.com/KobeissiLetter/status/1991255966235419112 ← +$205B market cap meme
https://x.com/amitisinvesting/status/1991263435493974047 ← Full breakdown thread
https://x.com/FromValue/status/1991275128439123451 ← “NVIDIA just printed more FCF than most companies make in revenue”
My Thoughts
This was a “relief rally on steroids”. Anyone still waiting for the AI capex slowdown just got obliterated. The $500 billion visibility isn’t hopium — it’s what they can already see in purchase orders.
The moat is now impenetrable: CUDA + NVLink + Spectrum-X + Grace CPU + Blackwell/Rubin roadmap = Microsoft Windows-level lock-in for the AI era.
At ~44× forward earnings the stock looks expensive until you realize the base case is now ~$260–280B annual revenue run-rate by late 2026. That puts the multiple in the low 20s. That is no longer the bull case — that’s the new floor.
The Christmas rally is officially back on. NVIDIA just saved it.
Bill Ackman’s Pershing Square Capital Management just released a 28-page investor presentation urging the Trump administration to immediately (1) deem the Treasury’s Senior Preferred Stock repaid, (2) exercise the 79.9% warrants, and (3) relist Fannie Mae (FNMA) and Freddie Mac (FMCC) on the NYSE — all while keeping the GSEs in conservatorship. They claim this can be done before the end of November 2025 and would instantly value the U.S. taxpayer’s stake at over $300 billion without disrupting mortgage affordability.
Key Takeaways
Fannie & Freddie OTC shares have already more than doubled in 2025 on Trump administration statements.
The three-step plan (repay SPS → exercise warrants → NYSE relisting) can be executed immediately by Treasury and FHFA.
Post-relisting, Treasury would own 79.9% of two NYSE-listed companies worth a combined ~$387 billion (Pershing estimate).
Taxpayers have already received $301 billion in dividends — $25 billion more than required under the original 10% deal.
Pershing strongly opposes any conversion of Senior Preferred into common — calls it value-destructive and legally risky.
Relisting unlocks massive institutional buying (many funds are barred from OTC stocks) and fulfills Trump’s campaign promise timing.
Conservatorship continues for years, giving the administration runway to finalize capital rules, backstop structure, and governance.
Detailed Summary of the Pershing Square Presentation (November 2025)
In a presentation titled “Promises Made, Promises Kept”, Pershing Square lays out a politically and financially attractive path for the second Trump administration to deliver on its GSE reform pledges without raising mortgage rates or rushing a full privatization.
The core argument: the government has already been fully repaid (and then some) via $301 billion of dividends since 2008. The Obama-era 2012 “Net Worth Sweep” was paused under Mnuchin, but never fully reversed. Pershing says a simple letter agreement between Treasury and FHFA can officially retire the Senior Preferred Stock today.
Once the SPS is gone, Treasury can exercise its long-held warrants for 79.9% of the common stock at essentially zero cost. The GSEs already meet every NYSE listing requirement (market cap, float, share price, shareholder count, etc.). FHFA can approve relisting while keeping full conservatorship powers intact — no change to operations, no new capital raises, no dividend payments to juniors until fully recapitalized.
Total taxpayer value: >$310 billion (plus junior preferred)
They explicitly reject the idea of converting Senior Preferred into common, warning it would trigger new litigation, force government consolidation onto the federal balance sheet, and slash valuations by 27–56% depending on the multiple the market would assign to a company that wiped out private shareholders.
My Thoughts
This is classic Ackman: aggressive, detailed, and perfectly timed to influence policy while he has a massive economic interest (Pershing owns large common positions in both GSEs). The beauty of the proposal is that it is genuinely low-risk from a mortgage-market standpoint and gives the administration an instant “win” before Thanksgiving 2025.
The politics line up perfectly: Trump gets to post on Truth Social that he turned two “bailed-out” companies into a $300 billion+ taxpayer windfall, keeps 30-year mortgage rates stable (or even lower), and still retains total control to shape the final exit over the next three years.
If Treasury and FHFA actually follow the three steps before November 30, 2025, the OTC-to-NYSE pop could be one of the largest wealth-transfer events in market history — and almost entirely to existing common shareholders (retail + hedge funds that held on since 2008).
Watch for any joint Treasury/FHFA announcement or letter agreement in the next two weeks. That will be the trigger.
Disclosure: Like Pershing Square, the author may have direct or indirect exposure to FNMA/FMCC securities.
Google just released Gemini 3 Pro – their smartest model ever. It crushes benchmarks in reasoning, coding, agentic workflows, and multimodal understanding. New tools include Google Antigravity (free agentic IDE), better bash/tool-calling, 1M context, and “vibe coding” that turns a single natural-language prompt or sketch into a full working app. Available today in Google AI Studio (free with limits) and via Gemini API at $2/$12 per million tokens.
Key Takeaways
Gemini 3 Pro is Google’s new flagship model (November 18, 2025) with state-of-the-art reasoning and agentic capabilities
Tops almost every major benchmark, including #1 on WebDev Arena (1487 Elo) and 54.2% on Terminal-Bench 2.0
New Google Antigravity – free public preview agentic development platform for Mac/Windows/Linux
1 million token context window + significantly better long-context usage than Gemini 2.5 Pro
Best-in-class multimodal: new SOTA on MMMU-Pro (image) and Video MMMU
Advanced “vibe coding”: build entire interactive apps/games from one prompt, voice note, or napkin sketch
New client-side & server-side bash tools, structured outputs + grounding, granular vision resolution control
Pricing (preview): $2/M input tokens, $12/M output tokens (≤200k context), free tiered after that
Free access (rate-limited) inside Google AI Studio right now
Already integrated into Cursor, Cline, JetBrains, Android Studio, GitHub, Emergent, OpusClip and many more
Detailed Summary of the Gemini 3 Launch
On November 18, 2025, Google officially introduced Gemini 3 Pro, calling it their “most intelligent model” to date. Built from the ground up for advanced reasoning and agentic behavior, it outperforms every previous Gemini version and sets new records across coding, multimodal, and general intelligence benchmarks.
Agentic Coding & Google Antigravity
The biggest highlight is the leap in agentic coding. Gemini 3 Pro scores 54.2% on Terminal-Bench 2.0 (vs 32.6% for Gemini 2.5 Pro) and handles complex, long-horizon tasks across entire codebases with far better context retention.
To showcase this, Google launched Google Antigravity – a brand-new, completely free agentic development platform (public preview for macOS, Windows, Linux). Developers act as architects while multiple autonomous agents work in parallel across editor, terminal, and browser, producing detailed artifacts and reports.
Vibe Coding & One-Prompt Apps
Gemini 3 Pro finally makes “vibe coding” real: describe an idea in plain English (or upload a sketch/voice note) and get a fully functional, interactive app in seconds. It currently sits at #1 on WebDev Arena with 1487 Elo. Google AI Studio’s new “Build mode” + “I’m feeling lucky” button lets anyone generate production-ready apps with almost zero code.
Multimodal Leadership
New SOTA on MMMU-Pro (complex image reasoning) and Video MMMU
Advanced document understanding far beyond OCR
Spatial reasoning for robotics, XR, autonomous vehicles
New client-side and hosted server-side bash tools for local/system automation
Grounding + URL context can now be combined with structured outputs
Granular control over vision fidelity (trade quality vs latency/cost)
New “thinking level” parameter and stricter thought-signature validation for reliable multi-turn reasoning
Pricing & Availability (as of Nov 18, 2025)
Gemini API (Google AI Studio & Vertex AI): $2 per million input tokens, $12 per million output tokens (prompts ≤200k tokens)
Free tier with rate limits in Google AI Studio
Immediate integration in Cursor, Cline, JetBrains, Android Studio, GitHub Copilot ecosystem, Emergent, OpusClip, etc.
My Thoughts
Gemini 3 Pro feels like the moment AI coding agents finally cross from “helpful assistant” to “can run an entire sprint by itself.” The combination of 1M context, 54% Terminal-Bench, and the new Antigravity IDE means developers can now delegate whole features or refactors to agents and actually trust the output.
The “vibe coding” demos (retro game from one prompt, full app from a hand-drawn sketch) are no longer parlor tricks – they are production-ready in Google AI Studio today. For indie hackers and prototyping teams this is an absolute game-changer.
Google pricing remains extremely aggressive ($2/$12) compared to some competitors, and giving Antigravity away for free is a bold move that will pull a huge portion of the agentic-dev-tool market toward their ecosystem overnight.
If you develop, design, or just have ideas – go download Antigravity and play with Gemini 3 Pro in AI Studio right now. 2026 is going to be built with this model.
Microsoft CEO Satya Nadella sat down with Stripe co-founder John Collison on the Cheeky Pint podcast in November 2025 for a wide-ranging, candid conversation about enterprise AI diffusion, data sovereignty, the durability of Excel, agentic commerce, and why today’s AI infrastructure build-out is fundamentally different from the 2000 dot-com bust.
TL;DW – The 2-Minute Version
AI is finally delivering “information at your fingertips” inside enterprises via Copilot + the Microsoft Graph
This CapEx cycle is supply-constrained, not demand-constrained – unlike the dark fiber of the dot-com era
Excel remains unbeatable because it is the world’s most approachable programming environment
Future of commerce = “agentic commerce” – Stripe + Microsoft are building the rails together
Company sovereignty in the AI age = your own continually-learning foundation model + memory + tools + entitlements
Satya “wanders the virtual corridors” of Teams channels instead of physical offices
Microsoft is deliberately open and modular again – echoing its 1980s DNA
Key Takeaways
Enterprise AI adoption is the fastest Microsoft has ever seen, but still early – most companies haven’t connected their full data graph yet
Data plumbing is finally happening because LLMs can make sense of messy, unstructured reality (not rigid schemas)
The killer app is “Deep Research inside the corporation” – Copilot on your full Microsoft 365 + ERP graph
We are in a supply-constrained GPU/power/shell boom, not a utilization bubble
Future UI = IDE-style “mission control” for thousands of agents (macro delegation + micro steering)
Agentic commerce will dominate discovery and directed search; only recurring staples remain untouched
Consumers will be loyal to AI brands/ensembles, not raw model IDs – defaults and trust matter hugely
Culture lesson: don’t let external memes (e.g. the “guns pointing inward” cartoon) define internal reality
Detailed Summary
The conversation opens with Nadella’s excitement for Microsoft Ignite 2025: the focus is no longer showing off someone else’s AI demo, but helping every enterprise build its own “AI factory.” The biggest bottleneck remains organizing the data layer so intelligence can actually be applied.
Copilot’s true power comes from grounding on the Microsoft Graph (email, docs, meetings, relationships) – something most companies still under-utilize. Retrieval, governance, and thick connectors to ERP systems are finally making the decades-old dream of “all your data at your fingertips” real.
Nadella reflects on Bill Gates’ 1990s obsession with “information management” and structured data, noting that deep neural networks unexpectedly solved the messiness problem that rigid schemas never could.
On bubbles: unlike the dark fiber overbuild of 2000, today Microsoft is sold out and struggling to add capacity fast enough. Demand is proven and immediate.
On the future of work: Nadella manages by “wandering Teams channels” rather than physical halls. He stays deeply connected to startups (he visited Stripe when it was tiny) because that’s where new workloads and aesthetics are born.
UI prediction: we’re moving toward personalized, generated IDEs for every profession – think “mission control” dashboards for orchestrating thousands of agents with micro-steering.
Excel’s immortality: it’s Turing-complete, instantly malleable, and the most approachable programming environment ever created.
Agentic commerce: Stripe and Microsoft are partnering to make every catalog queryable and purchasable by agents. Discovery and directed search will move almost entirely to conversational/AI interfaces.
Company sovereignty in the AI era: the new moat is your own fine-tuned foundation model (or LoRA layer) that continually learns your tacit knowledge, combined with memory, entitlements, and tool use that stay outside the base model.
Microsoft’s AI stack strategy: deliberately modular (infra, agent platform, horizontal & vertical Copilots) so customers can enter at any layer while still benefiting from integration when they want it.
My Thoughts
Two things struck me hardest:
Nadella is remarkably calm for someone steering a $3T+ company through the biggest platform shift in decades. There’s no triumphalism – just relentless focus on distribution inside enterprises and solving the boring data plumbing.
He genuinely believes the proprietary vs open debate is repeating: just as AOL/MSN lost to the open web only for Google/Facebook/App Stores to become new gatekeepers, today’s “open” foundation models will quickly sprout proprietary organizing layers (chat front-ends, agent marketplaces, vertical Copilots). The power accrues to whoever builds the best ensemble + tools + memory stack, not the raw parameter count.
If he’s right, the winners of this cycle will be the companies that ship useful agents fastest – not necessarily the ones with the biggest training clusters. That’s excellent news for Stripe, Microsoft, and any founder-focused company that can move quickly.
FINAL UPDATE – Post-Mortem Released: Cloudflare has released the detailed post-mortem for the November 18 event. The outage was caused by an internal software error triggered by a database permission change, not a cyberattack[cite: 25, 26]. Below is the technical breakdown of exactly what went wrong.
TL;DR – The Summary
Start Time: 11:20 UTC – Significant traffic delivery failures began immediately following a database update.
The Root Cause: A permission change to a ClickHouse database caused a “feature file” (used for Bot Management) to double in size due to duplicate rows[cite: 26, 27, 81].
The Failure: The file grew beyond a hard-coded limit (200 features) in the new “FL2” proxy engine, causing the Rust-based code to crash (panic)[cite: 190, 191, 194].
Resolution: 17:06 UTC – All systems fully restored (Main traffic recovered by 14:30 UTC)[cite: 32, 90].
The Technical Details: A “Panic” in the Proxy
The outage was a classic “cascading failure” scenario. Here is the simplified chain of events from the report:
The Trigger (11:05 UTC): Engineers applied a permission change to a ClickHouse database cluster to improve security. This inadvertently caused a query to return duplicate rows[cite: 160, 172].
The Bloat: This bad data flowed into a configuration file used by the Bot Management system, causing it to exceed its expected size[cite: 27, 125].
The Crash: Cloudflare’s proxy software (specifically the FL2 engine written in Rust) had a memory preallocation limit of 200 features. When the bloated file hit this limit, the code triggered a panic (specifically called Result::unwrap() on an Err value), causing the service to fail with HTTP 500 errors[cite: 190, 218, 219].
The Confusion: To make matters worse, Cloudflare’s external Status Page also went down (returning 504 Gateway Timeouts) due to a coincidence, leading engineers to initially suspect a massive coordinated cyberattack.
Official Timeline (UTC)
Time (UTC)
Status
Event Description
17:06
Resolved
All services resolved. Remaining long-tail services restarted and full operations restored[cite: 268].
14:30
Remediating
Main impact resolved. A known-good configuration file was manually deployed; core traffic began flowing normally [cite: 32, 268].
13:37
Identified
Engineers identified the Bot Management file as the trigger and stopped the automatic propagation of the bad file [cite: 268].
13:05
Mitigating
A bypass was implemented for Workers KV and Access to route around the failing proxy engine, reducing error rates [cite: 267].
11:20
Outage Starts
Network begins experiencing significant failures to deliver core traffic .
11:05
Trigger
Database access control change deployed[cite: 267].
Final Thoughts
Cloudflare’s CEO Matthew Prince was direct in the post-mortem: “We know we let you down today”[cite: 37]. The company has identified the specific code path that failed and is implementing “global kill switches” for features to prevent a single configuration file from taking down the network in the future[cite: 259].
November 18, 2025 – Bitcoin just printed $89,420, its lowest level since February. That single candle erased every bit of 2025’s gains and pushed the price down more than 29% from the October all-time high of $126,250.
The Crypto Fear & Greed Index is sitting at 11 – the lowest since the 2022 bear market. We are officially in “extreme fear” territory.
This is not a crypto-specific meltdown. BTC is simply doing what it always does in risk-off environments: acting like the most leveraged tech stock on earth.
What Actually Broke the Market
A confirmed death cross (50-day MA under 200-day MA) combined with a clean break of the 200-day moving average.
U.S. spot Bitcoin ETF inflows have gone from +$25B earlier this year to dead flat / negative for almost two weeks straight. BlackRock’s IBIT alone saw a record ~$1.26B outflow this month.
Corporate treasury buyers (MicroStrategy-style) have hit the pause button.
OG whales who stacked pre-2024 continue distributing at the fastest pace of the cycle.
Newer buyers (last 12–24 months) are panic-selling 20–50% gains because “the bull market peaked in 2025” FUD is everywhere.
“The recent sell off is from newish buyers… They bought in the last 12-18 months… and are ‘taking profit’ for 20%-30% fiat gains… This cohort of sellers is also depleted, and HODLers with conviction have now taken their coins.”
Sentiment this bad almost never lasts. On-chain data shows long-term holders are absorbing supply, Bitcoin dominance is spiking (alts getting wrecked first), and the Fear & Greed Index below 15 in a post-halving bull year has always marked monster reversals.
Top Bottom-Calling Tweets Right Now
“Bitcoin is literally free at $91,000. The bottom is near, and then BTC will pump straight to $150,000!!”
“I’ve only seen sentiment this bad 3 times before: 2018 lows, COVID crash, FTX collapse. Things can stay negative for awhile, but I find it hard to believe we’re closer to the top than to the bottom.”
On November 17, 2025, Valar Atomics and Los Alamos National Laboratory announced that the NOVA Core – a HALEU TRISO-fueled, graphite-moderated HTGR test assembly – successfully reached zero-power (“cold”) criticality at the National Criticality Experiments Research Center (NCERC) in Nevada. This marks the first time a venture-backed private nuclear company has ever achieved criticality, validating the physics of Valar’s upcoming Ward250 reactor and clearing a major technical de-risking milestone on the path to gigawatt-scale carbon-free power.
Key Takeaways
Zero-power criticality achieved at 11:45 AM PT on November 17, 2025
First criticality ever achieved by a venture-funded nuclear startup
Conducted at the United States’ only general-purpose critical experiments facility (NCERC, Nevada National Security Site)
Uses the exact same HALEU TRISO fuel, graphite moderator, and reactivity control scheme as the commercial Ward250 reactor
Directly validates Valar Atomics’ proprietary neutronics models and simulation stack
Builds on the 2024 Deimos critical assembly; NOVA is the high-fidelity physics twin of Ward250
Clears the path for hot (powered) criticality and full-temperature testing in 2026
Supported by DOE’s Advanced Reactor Pilot Program (target: full criticality by July 4, 2026) and Executive Order 14301
Strong public endorsement of the Trump administration’s “make nuclear great again” push
Detailed Summary of the Announcement
On November 17, 2025, Los Alamos National Laboratory (LANL) and Valar Atomics jointly announced that the NOVA Core, operating on LANL’s Comet critical assembly machine at the National Criticality Experiments Research Center (NCERC) inside the Nevada National Security Site (NNSS), had achieved zero-power criticality at exactly 11:45 AM Pacific Time.
Approach-to-critical experiments began on November 12, 2025, and the core went critical five days later – an impressively rapid and safe execution that highlights both Valar’s engineering maturity and NCERC’s world-class operational capability.
What is zero-power (“cold”) criticality?
Cold criticality is the moment when a nuclear core sustains a stable neutron chain reaction (k_eff = 1.000) without external neutron sources, but at room temperature and with essentially zero fission power (typically microwatts to a few watts). No heat is removed by coolant flow, and temperatures remain ambient. It is the nuclear equivalent of “first breath” or “first heartbeat” – proof that the fundamental physics of the core design works exactly as modeled.
Project NOVA (Nuclear Observations of Valar Atomics) is a multi-week campaign of criticality experiments designed to:
Measure integral neutronics parameters (reactivity coefficients, control rod worth, burnable poison performance, etc.)
Validate Valar’s in-house Monte Carlo and deterministic neutronics codes
Provide high-fidelity benchmark data for the Ward250 reactor currently under construction in Utah
The NOVA Core is a graphite-moderated, helium-cooled-concept test bed fueled with High-Assay Low-Enriched Uranium (HALEU) TRISO particles – the same fuel form and enrichment Valar will use commercially. Reactivity control is provided by boron-carbide elements in stainless-steel cladding, mirroring the Ward250 design.
The central portion of the core was designed and fabricated entirely by Valar Atomics, while LANL provided the Comet universal assembly machine, reflectors, instrumentation, safety envelope, and decades of criticality-safety expertise.
Quotes from Leadership
Isaiah Taylor (Founder & CEO, Valar Atomics): “Zero power criticality is a reactor’s first heartbeat, proof the physics holds… This moment marks the dawn of a new era in American nuclear engineering — one defined by speed, scale, and private-sector execution with closer federal partnership.”
Max Ukropina (Head of Projects): “President Trump asked industry and the labs to make nuclear great again. We got together and decided to start with the basics of fission. This team delivered incredible results safely so we can keep moving up the technical ladder.”
Sonat Sen (Lead Core Designer): “Project NOVA provides us with real-world data which will help us answer key questions about TRISO fuel performance in our core and validate our proprietary software stack.”
Why This Milestone Matters – Technical & Strategic Context
Reaching criticality in a national-lab critical facility is widely regarded as the single biggest technical de-risking event for any new reactor design. Before today, no venture-backed nuclear company had ever achieved criticality on their own core. Legacy players (NuScale, TerraPower, Kairos Power, X-energy, etc.) have either used legacy government assemblies or have not yet gone critical with their exact commercial fuel and geometry.
Valar Atomics has now leapfrogged the field by:
Using actual commercial-spec HALEU TRISO (not surrogates)
Replicating the exact Ward250 moderator-to-fuel ratio and control scheme
Collecting integral data months ahead of first fuel load at Ward250
Demonstrating that a small private team can execute at national-lab speed and safety standards
This positions Valar to move aggressively into hot zero-power testing, helium loop commissioning, and ultimately full-power, full-temperature operation of Ward250 in 2026 – aligning perfectly with the DOE’s goal of new reactor criticality by Independence Day 2026.
My Thoughts & Broader Implications
1. Speed is the new moat. From Deimos (2024) → NOVA criticality (2025) → Ward250 power operations (2026) in roughly 24 months is an absolutely blistering pace by historical nuclear standards. Valar is proving that private capital + national lab partnership + focused scope can compress decades into years.
2. TRISO + Graphite + Helium is having its moment. The combination of walk-away-safe TRISO fuel, high-temperature capability (>750°C), and modular factory fabrication is rapidly becoming the consensus Gen-IV architecture for private deployment. NOVA just added the strongest data point yet that the neutronics actually work as advertised.
3. National labs are back as force multipliers. NCERC’s ability to take a private core, insert it into the Comet machine, and go critical in under a week with zero safety incidents is a national strategic asset. The close LANL–Valar collaboration is exactly the model the Trump administration appears to want: labs providing capability, private sector providing speed and capital.
4. AI + Nuclear inflection point. Valar has been explicit that their ultimate product is gigasites – clusters of thousands of HTGRs powering hyperscale data centers, hydrogen electrolysis, and desalination. Today’s criticality is concrete evidence that the energy bottleneck for the AI build-out may actually be solvable in this decade.
5. First of many. If Valar can replicate this model – design core → validate at NCERC → deploy Ward250 → scale factory production – we are looking at a genuine nuclear renaissance led by American startups rather than slow-moving utilities or foreign state-owned entities.
Wrap Up
November 17, 2025, will be remembered as the day a venture-backed nuclear company first split the atom under its own design. Project NOVA’s successful cold criticality is not just a technical checkbox – it is a cultural and strategic turning point for the entire industry.
The physics works. The team can execute. The labs are partnering at speed. The policy tailwinds are strong.
We are witnessing the birth of the next era of American nuclear dominance – and it’s moving a lot faster than anyone predicted.
