Jeff Dean

In April, '17, @jsomers of @NewYorker reached out & said he wanted to do a small profile of me & my longtime colleague Sanjay Ghemawat, watch us work for a few hours, maybe dinner, etc. It came out today. I think it captures our working style really well. newyorker.com/magazine/2018/…

New paper: every law in America is technically public. But not really, until now! With @DenisPeskoff at UC Berkeley, we built a corpus of ~every publicly accessibly city and county law, and released a huge chunk of it! 2.2 million laws, you're (probably) covered in it! 🧵

@JohnJumperSci @demishassabis All the best, John! Every time our paths have crossed I have deeply enjoyed our interactions!

A bit of news: After nearly 9 years, I have decided to leave Google DeepMind and join Anthropic (after taking some time to recharge). I am incredibly grateful for my time at GDM. @demishassabis took a real chance letting me lead the AlphaFold team just six months after finishing my PhD, and the entire GDM team taught me so much about how to do great science. GDM is a special place, and I’ll still be excited to hear about what amazing things they discover next.

@sujay_kapadnis @arcjax7 Because I didn't have more time to chat with @arcjax7!

Today, @aadityasubedi_ and I are excited to introduce @architectlabs. We are building the AI system to design and provably verify chips for the world's most demanding workloads. AI scaling is fundamentally changing the economics of hardware infrastructure. As models scale, become more capable, and more widely deployed, the bottleneck is shifting from software and models alone to the physical infrastructure that runs it: specialized compute, memory, networking, interconnects, and full-stack system design. General-purpose hardware is no longer enough, and the world is racing to spin up new chip programs. This is not just true across datacenter training and inference, but everywhere AI enters the physical world: robotics, autonomous systems, spatial computing, defense, personal devices, industrial automation, and scientific instruments. But designing a chip today remains one of the most gated efforts in modern technology. A modern chip program takes years, costs hundreds of millions of dollars, and depends on a shrinking pool of expertise concentrated inside a small number of companies. Architect Labs is a foundational lab building an AI system that designs and provably verifies chips end-to-end. We partner with semiconductor and workload companies, AI labs, and nations to turn demanding workloads into purpose-built chips, on demand at scale. We aim to drastically accelerate chip design, so that the models, software and chip designs can co-evolve together, accelerating the industry’s path to superintelligence. Two decades ago, the fabless revolution made it possible to build a chip company without owning a fab. TSMC made world-class manufacturing available to anyone with a design. We aim to do the same for chip design itself: enable any organization with a workload, or specification, to get a purpose-built chip design that unlocks scale and distribution of intelligence impossible with current hardware paradigms. Our founding team collectively has taped out 80+ production chips, led $10B datacenter product lines, been core contributors to Meta’s AI silicon, architected and designed one of the first neuromorphic chips out of Intel, led research teams at Anthropic, xAI, and Google DeepMind, and contributed to fundamental AI research across nearly every frontier lab. We have already partnered with semiconductor companies to accelerate their chip programs, and some of our AI-generated chip designs are going to tape out on leading-edge foundry nodes later this year. We’ve also raised a $24M seed round led by @stevejang from @KindredVentures , with participation from @TQVentures , @RaceCapital , @scaletogether , @ora, and Link Ventures. We are grateful for the support of our angels and advisors, including @snsf , @lukaszkaiser , @AravSrinivas , Kunle Olukotun, @tlbtlbtlb, @alexwg, Siddharth Nath, Thierry Tambe, @arashf , @ekaurghar, @CHHubbell, Selene Casabal, @semiDL , and engineering leaders from OpenAI, NVIDIA, Google DeepMind, Intel and more. The next great scaling law may not come from the models alone. It will come from making the physical substrate of intelligence programmable. We exist to bring this future to life.

@axi_master @aadityasubedi_ @architectlabs Congrats!

Stephen Colbert is wearing a tan suit at the Obama Presidential Center grand opening! 🤣💙

@NoamShazeer All the best in everything you do, Noam! I've enjoyed the work we've done together!

I’m excited to share that I’ll be joining OpenAI and look forward to working with the exceptional team there. It was a difficult decision to move on. I’m incredibly proud of the amazing team at Google and everything we’ve built together. It has been an honor and a pleasure to work with all of you.

Everyone says the latest AI agents will be "job-ready" soon, especially after the release of Fable 5 this week. But is that really the case? Over the past many months, my group and collaborators have been building Agents' Last Exam (ALE), a benchmark designed to test exactly that claim on real digital labor-market work. My group and collaborators previously have created many of the benchmarks the field runs on, including MMLU, MATH, CyberGym, and ExploitGym. Today, I'm excited to share Agents' Last Exam (ALE): a rolling benchmark that measures whether AI agents can actually perform economically valuable work across a broad range of real-world domains. With ALE, we evaluated Fable 5, GPT-5.5, Composer 2.5, and other frontier agent systems across more than 1,500 expert-sourced tasks spanning 55 occupations. The result is both impressive and sobering. Today's agents can solve a meaningful fraction of professional tasks. But when we look at the hardest tasks, the ones requiring sustained reasoning, deep domain expertise, and reliable execution over long horizons, they are still far from human-level performance. On ALE's hardest tier, every frontier agent we tested, including Fable 5, achieved a 0% success rate. The age of useful agents is here. The age of truly job-ready agents is not. We hope Agents' Last Exam (ALE) will serve as a new guidepost and north star for developing agents capable of reliably performing economically valuable work across a broad range of domains. 🧵

My @Google colleagues @NormJouppi, Sridhar Lakshmanamurthy, Cliff Young, and David Patterson recently wrote a paper that will appear in the July/August 2026 edition of @ieeemicro titled "Google's Training Supercomputers from TPU v2 to Ironwood: Architectural Stability, Scale, Resilience, Power Efficiency, and Sustainability Across Five Generations". It's chock full of interesting data about the evolution of TPU chip generations, as well as how workloads at Google have transformed over time (hint: lots more transformer-based models!), and how the generations have gotten ~30X more energy efficient per flop. Lots of changes over these generations: Air cooling in TPUv2 to water cooling in TPUv3 onwards 2D to 3D torus-based interconnects 30X improvement TFLOPS/Watt 256 chips (TPUv2) to 9216 chips (Ironwood) per pod Read the full paper: arxiv.org/abs/2606.15870

Together with my co-founders Michael @MichaelPoli6, Stefano @Massastrello and Armin @athmsx, I am excited to announce @RadicalNumerics is emerging from stealth with a $50M seed round to build general biological intelligence. We’re also sharing an early preview of our new model Omnii, the most powerful genome language model to date. Omnii preview link: radicalnumerics.ai/blog/radical-n… At Radical Numerics, our mission is to master the code of life, and to drive the frontier of biological AI for both design and defense. This is our dual mandate, which comes from something our own team helped make possible. Our founding team trained Evo and Evo 2, the largest biological AI models (40B params) trained on DNA sequences. Trillions of tokens across all of life, from microbes to mammals. It’s fully open source, and created the field now known as generative genomics. Last year, scientists used Evo to generate the world’s first complete genome from scratch using AI. Turns out it was a bacteriophage—a type of virus. It functioned in the real world, and in this case it was harmless. But for us, it was a clear turning point. It showed that AI is no longer just analyzing biology. It is on the cusp of generating functional lifeforms. Eventually, AI will have the power to design and control life itself. That should make all of us incredibly excited, and incredibly uneasy. (Anyone can design DNA with a new function, and have it synthesized and delivered, like something from Amazon Prime). The same technology that will help us cure cancer is the very technology that might create the next global pandemic, or worse, allow the creation of bioweapons that can wipe out populations. We believe these forces are inseparable. If you work on the frontier of biology, you have to build technology to safeguard it from its misuse. Existing biosecurity tools are sorely losing the arms race, relying on outdated “have I seen this exact thing before?” style algorithms. We founded Radical Numerics to turn the tide. And we can’t do that by training on textbooks and natural language. We must understand the language of biology from the raw physical data itself, to reason across every molecule and modality, from DNA to proteins. The next frontier for AI goes far beyond chatbots or video generators to models that can understand and engineer life. Today, we’re previewing Omnii, which is already far surpassing Evo 2, and will continue improving as we scale and add new modalities (training now). 1. For human health, Omnii can read and write whole genomes (more on writing later). It’s state of the art (SOTA) on detecting causal variants for disease, and can rank Alzheimer's mutations zero-shot. We’re partnering with a diagnostics company to use Omnii for early cancer detection (pancreatic and multi-cancer). 2. For defense, Omnii is SOTA at detecting AI-generated pathogens. We benchmarked existing detection tools, and they simply can’t detect the AI-generated ones (“deepfake viruses”). We’re partnering with a US national lab to pilot Omnii for detecting the next pandemic, both natural and AI-generated. We have a data center full of Blackwells in construction now to build the most powerful biological AI models ever. This mission takes a new kind of AI lab that can actually scale on physical, biological data: new alignment research (mid/post training), scaling long context, building out mech interp teams to dissect what these models learn, new architectures and systems designs, all from the ground up. Our team is made up of AI researchers and scientists from top labs and institutions (e.g. Stanford, MIT, Google DeepMind), but more importantly, we all share the belief that this is the most important challenge of our lifetime. If you feel similarly, we are hiring. We aim to bring the brightest minds in AI and science together to save lives. Thanks to our partners on this journey, led by Emergence Capital @emergencecap, with Obvious Ventures @obviousvc, Triatomic @TriatomicCap , and Patrick Collison @patrickc. Our advisors include Eric Horvitz @erichorvitz, CSO of Microsoft, Chris Re @HazyResearch of Stanford, George Church @geochurch of Harvard, and Andrew Weber @AndyWeberNCB, former Assistant Secretary of Defense for Nuclear, Chemical and Biological Defense Programs. Fortune article: fortune.com/2026/06/15/exc… Jobs: radicalnumerics.ai/join-us

@achettimada That game was also great!

@JeffDean Mr. Dean, Iran vs New Zealand was much more exciting.

I really enjoyed this game today. Vozinha and the whole Cape Verde team were amazing against Spain!

A good essay by @pgasawa and @profjoeyg on a more nuanced view of AI advances.

The AI community seems to increasingly be heading towards a polarized world when discussing safety and consolidated power. I see this discourse as a false dichotomy, so @profjoeyg and I wrote an essay on how we need to change the conversation (link below).

@martin_peris @Google @UCSD Sorry, I phoned it in.

@JeffDean @Google @UCSD No dad joke?? I was waiting for the punch line 😜

People replace their phones every ~4 yrs. This means there are hundreds of millions of old phones discarded each year that are still perfectly usable as computing devices. @Google in collabration with @UCSD is exploring how to turn these old phones into cloud-computing “phone clusters”. Putting phones back in service in this way can directly reduce the environmental footprint of computing by avoiding the need for further raw material extraction, and taking advantage of the embodied carbon already incurred from manufacturing these devices, and modern phones actually are already quite powerful computers. Read more in the blog below ⬇️

@uwcse @MBalazinska Here's a nice @GeekWire article by @ToddBishop about the graduation, including a few more snippets of what I said in my talk. geekwire.com/2026/what-a-lo…

@uwcse @MBalazinska Here are a few of the pieces of advice that I shared:

I enjoyed giving the commencement address at the University of Washington Allen School @uwcse graduation this evening. So many happy students and their families and friends! Congratulations to all the graduates of the class of 2026! 🎓 Thanks for inviting me, @MBalazinska!

@CashMoneyLemon @uwcse @MBalazinska Nope! I offered an optimistic view of the great potential of AI, but also a realistic view that we need safeguards and boundaries to ensure that we maximize the benefits and minimize the downsides.

@JeffDean @uwcse @MBalazinska Any AI boos?

@MacroBombastic @Google @UCSD We remove the batteries. x.com/i/status/20656…

@JeffDean @Google @UCSD Interesting angle but running a cluster of old batteries sounds like a fire hazard waiting to happen mate