


Backing inventor-founders building strange, ambitious things. I write and build to think thereview.strangevc.com



Everything is breaking. Nov 17: Apple’s iCloud, App Store and Apple Pay go down; users are signed out of their accounts. No cause given. Nov 17: the NSA’s classified network goes down. No cause given. Nov 16: a software bug at Chase stalls payments for hours, including transfers to customers at other banks. Nov 16: hackers use a flaw in Optimizely to plant card-skimming code on 178 company websites, including Zoom, CNN and HSBC. Nov 15: a Verizon outage cuts 911 calls across 13 Pennsylvania counties. No cause given. Nov 13: Workday goes down. A test of its backup systems corrupted live data. Nov 13: the FAA grounds every US flight for almost 5 hours. A router installed during a network upgrade flooded the system. Nov 12: Microsoft 365 goes down in 70+ countries. A server change from July broke under extra load.

Jensen Huang didn't just build Nvidia. He helped build the market for AI compute on both the demand and supply sides. For decades, he cultivated suppliers like TSMC, secured manufacturing capacity, and convinced customers to invest billions in AI infrastructure. Today, Nvidia is a buyer, supplier, investor and guarantor across nearly every layer of the compute stack. Now Jensen is out after 33 years. Can the extraordinary web he built hold together without him? I mapped the relationships (and the billions behind them) in this week's Strange Brief. Also this week: → SpaceX catches a Starship returning from orbit → Google releases Gemini 4 and new agent orchestration tools → AI mathematically verifies security properties of Signal's code → Microsoft's global outage exposes infrastructure dependencies → Advanced nuclear moves closer to the factory floor Read on for your weekly catchup on the news shaping compute: open.substack.com/pub/strang…

This is an absolute black swan event. Nvidia without its founder is a lame duck. Jensen was constantly pushing the ecosystem forward both here and in our manufacturing partners in Taiwan. Expecting major shake ups across the demand and supply side of compute over the next years once we start seeing the effects of that gone.

Apple's darkhorse AI strategy is starting to come into focus. In the last few days: • A new leak suggests Harbor could help you run open models on a cluster of Macs. • OpenAI’s dots are going multi-computer, joining the DIY crowd already running agents like Hermes and OpenClaw on Mac minis. • Adobe is opening its creative tools to third-party AI agents through upcoming MCPs. You can see where this is going. The Mac is becoming a place where models can run and agents can work, with your apps, files, and accounts already there. Apple’s bet may be that the last mile of AI happens on-device and with the iOS ecosystem. Today, you buy a Mac for yourself. Tomorrow, you might buy a bunch more to run for your agents and models. Read on The Strange Review: open.substack.com/pub/strang…

What a week. Tech stocks bounced somewhat back post midterms on Friday, just as fresh details emerged about an alleged Chinese intelligence operation targeting Nvidia's CEO. He has now been asked to brief congressional committees. No evidence of wrongdoing. Timing is odd. But it makes you wonder about the amount of concentration risk we have in the AI boom. Also this week: → Marvell's $115B bet on optical connectivity → MIT researchers use AI to rewrite another model's weights → An autonomous biology lab runs 11,000 experimental steps → OpenAI uses AI to uncover hundreds of Linux vulnerabilities → Engineers reverse-engineer a GPU megakernel from a video All in this week's The Strange Review, a Sunday recap on the news in compute this week. open.substack.com/pub/strang…

The scariest thing in tech this Halloween for our readers might be the market correction this week. With midterms approaching and AI spending under scrutiny, the Nasdaq fell ~5% in October. Alphabet lost 12.6% in a day. And yet AI demand is booming. Anthropic reported $16.3B in six-month revenue. Google's cloud business doubled. So what's spooking investors apart from election jitters? Perhaps the $202B that Amazon, Google, Microsoft and Meta spent on infrastructure in just ONE quarter. More than they spent in all of 2022. The last time we saw an infrastructure frenzy like this was during the dot-com boom. I dug into the numbers, with five charts tracing today's AI buildout. Also this week: → Google's Genie 4 brings interactive world models closer. → Wall Street arranges $63B to finance Anthropic's AI chips. → Google runs an AI model on a TPU in space. → OpenAI claims a breakthrough in computational complexity. Join the investors and founders reading this before Monday, in this week's Strange Brief ↓ open.substack.com/pub/strang…

Robots making medicine. A new path to fusion. And plenty of good questions about what comes next. A lovely write-up of Strange Summit in @Inc! Thanks to Dave Benton for joining us and capturing the spirit of the day. The big takeaway: it isn't just about what we can build, but what's worth building. Read on ↓ inc.com/dave-benton/i-went-l…

Everyone's backup business in AI seems to be selling compute. GPU rental prices are rising. Inference providers are popping up. And some AI labs are treating compute reservations as a fallback business. A quick read before Monday to stay up to date on what happened this week: → OpenAI's agents escape security sandboxes → Anthropic generates formal software proofs → DeepSeek opens its inference stack → Google and Amazon challenge Nvidia's software moat → Microsoft pulls out of a data-center deal over power delays → Nvidia's robots, Moonshot's new model, Starlink and more Plus: Apollo's chief economist questions whether AI has paid off outside Big Tech. Read in The Strange Review ↓ open.substack.com/pub/strang…

Anthropic's numbers are extraordinary. - $4.6B revenue in all of 2025 - $16.8B in Q3 2026 alone - 45% Q-on-Q growth (down from 142%, but calling that disappointing at this scale seems slightly absurd) - Compute spending has fallen from $1.60 of revenye to roughly $0.59 per dollar of revenue The business is growing at an astonishing rate, and its economics are improving dramatically. But Anthropic has also committed $518B to future computing infrastructure to compete in a market where a model's lead might last only months. A few months ago, I wrote about the finite game of frontier AI. Anthropic's financials give a real sense of how the economics are playing out (link below for some easy-to-read chart). Also this week: → China is doubling down on semiconductor manufacturing. A new tracker maps 57 facilities, and capacity is projected to grow 8.1% annually, versus 5.3% globally. Could chips follow the path of solar panels and batteries? → AI infrastructure stocks take a beating. Even though demand for compute is exploding, neoclouds are billions in debt. With borrowing costs rising, investors are getting nervous about who will finance the AI buildout. → Meta's new Muse Wonder models and Google's Gemini 4 Argon releases this week are pushing AI toward agents that remember, reason and work through longer tasks. → AI learns from its mistakes in the lab. Google's agents used failed physical experiments to improve molecular-design hit rates from 3.7% to 33.5%. Read on for my analysis of the news that moved the world of intelligence this week in The Strange Brief ↓ open.substack.com/pub/strang…

My hunch is that we'll each end up with one primary AI agent, not five. Personal agents could have some of the strongest lock-in effects we've ever seen in software. Think Apple vs. Android, except the switching costs might be even higher. An agent that knows your work, your relationships, your preferences, your quirks becomes harder to replace with every interaction. I've been working with OpenAI's Dots lately, and it's the first personal assistant that made this feel real. It's incredibly proactive. And creative. It found solutions and workarounds to problems I was grappling with, without me even explicitly asking. There might be plenty of specialized tools, but one agent that knows you and coordinates the rest. The race to acquire users will be fierce. Getting them to switch might be even harder. The window to gain ground is now. Also this week: → SpaceX is taking on AT&T, Verizon and T-Mobile. The three lost $62B in market value following its spectrum deal. → OpenAI released 722 AI-generated math manuscripts, then withdrew three the next day. Who checks all this work? → Google is adding 890 MW of nuclear power by upgrading existing reactors. No new reactors needed. → GlobalFoundries signed a $2B deal with TSMC as advanced chip packaging becomes an AI bottleneck. → Lumentum says its optical-component capacity is nearly sold out through 2029. Connecting GPUs is becoming a bottleneck of its own. → DeepSeek is reportedly raising at least $12B, backed by Tencent. More in this week's The Strange Brief, where I unpack the ideas and inventions changing the future of computing ↓ open.substack.com/pub/strang…



More than $11B went into AI drug discovery last year. But finding a new medicine is only useful if you can actually make it. Some of the most advanced therapies in the world are still produced through painstaking manual work inside cleanrooms. Extraordinary science running on surprisingly manual infrastructure. That is the problem Multiply Labs is going after. They are building the factory infrastructure for the next generation of medicine: robotic systems that can take on the delicate, repetitive work behind cell therapy manufacturing and more. Today, the manual cell therapy process produce an average annual throughput of roughly 10,000 doses. The waitlist for the medicine is orders of magnitude more. As medicine gets more complex, manufacturing becomes one of the biggest bottlenecks. The factory starts to matter almost as much as the drug. Congrats Multiply Labs on the Series B! Excited to be (finally) on the journey. Read on for why we invested here: open.substack.com/pub/strang…

The venture market feels a little unhinged right now. - bitchfights between investors on behalf of port cos - valuations multiplying like crazy on the smallest whiff of hype In the meantime, some of the hottest new entrants have impressive copycats from startups and big companies within days, if not weeks. Is there a moat? Maybe. Maybe no one cares. Right now the market seems perfectly happy to jump in first and figure that part out later. Also this week: → Anthropic’s leaked IPO prospectus shows what it costs to run a frontier lab, including $518B in future cloud, compute and infrastructure commitments → OpenAI used DevDay to fill in more of the agent stack: hosted execution, memory, computer use, multi-agent orchestration and a new Decisions API → DeepMind can now watermark AI-generated proteins in a way that survives physical synthesis → “Decision models” went from Jev to Clef, Jeff, Jeeves and OpenAI’s Decisions API in roughly two weeks → AMD is buying World Labs for $8.2B and bringing frontier-model research closer to the hardware roadmap → OpenAI and Synopsys are building a specialized model for chip design → Apple is tightening Full Disk Access as autonomous agents make old permission models look increasingly risky Get it all in The Brief: **The Market Moves Fast** ↓ open.substack.com/pub/strang…