I am very impressed with @Google's new Gemini 3 AI model which was released yesterday. From my early testing, Gemini 3 seems like it is the best frontier AI model out there, besting GPT-5 from @OpenAI and Claude from @Anthropic on my private evaluations. gemini.google.com/app

‎Google Gemini

Meet Gemini, Google’s AI assistant. Get help with writing, planning, brainstorming, and more. Experience the power of generative AI.

gemini.google.com

In celebration of Stanford's Codex Center's 20 year anniversary, Stanford Law School produced this retrospective video about the center's founding and work. I am grateful to Roland Vogl, Mike Genesereth, Megan Ma, and the rest of the amazing Community. www.youtube.com/watch?v=Bc4V...

In celebration of Stanford's Codex Center's 20 year anniversary, Stanford Law School produced this retrospective video about the center's founding and work. I am grateful to Roland Vogl, Mike Genesereth, Megan Ma, and the rest of the amazing Community. www.youtube.com/watch?v=Bc4V...

My keynote address on "Artificial Intelligence, Autonomous Systems, and Law" from last week's Silicon Flatirons Conference at the University of Colorado law school, is now available online for those interested. youtu.be/OX7hfzBT40c?...

Artificial Intelligence, Autonomous Systems, and Law | Welcome and Keynote

YouTube video by Silicon Flatirons

youtu.be

Join us this Friday March 7 for a conference on "Autonomous AI and Law" This conference explores the novel law of emerging AI agents like @OpenAI Operator that decide their own actions based on high-level goals. cc:@ColoLaw @SiliconFlatiron siliconflatirons.org/events/2025-...

@OpenAI released their latest AI model GPT-4.5 yesterday (a research preview). So far, I am *not* very impressed. It definitely doesn't seem meaningfully "smarter" than GPT-4 or other earlier frontier models on my internal tests, despite being a larger model. openai.com/index/introd...

New paper: Simulating Time With Square-Root Space people.csail.mit.edu/rrw/time-vs-... It's still hard for me to believe it myself, but I seem to have shown that TIME[t] is contained in SPACE[sqrt{t log t}]. To appear in STOC. Comments are very welcome!

Please join me on March 7, 2025 for a conference I am hosting on "Artificial Intelligence, Autonomous Systems, and Law" at Colorado Law's Silicon Flatirons Center. siliconflatirons.org/events/2025-... This is among the first law school conferences studying advances in autonomous AI agents.

Please join me on March 7, 2025 for a conference I am hosting on "Artificial Intelligence, Autonomous Systems, and Law" at Colorado Law's Silicon Flatirons Center. siliconflatirons.org/events/2025-... This is among the first law school conferences studying advances in autonomous AI agents.

This is a great talk by the always terrific Stanford University Professor Erik Brynjolfsson on the impact of artificial intelligence on the workforce at the 2025 Paris AI Action Summit. youtu.be/6HDjVncL5Go?...

AI, Science and Society Conference - AI ACTION SUMMIT - DAY 2

YouTube video by IP Paris

youtu.be

Join us this Sunday 2/2 and Monday 2/3 for our Silicon Flatirons flagship conference at University of Colorado Law School on "Examining The New Federalism in Technology Policy" I will be on a panel on AI regulation at the state level. siliconflatirons.org/events/2025-...

Join us this Sunday 2/2 and Monday 2/3 for our Silicon Flatirons flagship conference at University of Colorado Law School on "Examining The New Federalism in Technology Policy" I will be on a panel on AI regulation at the state level. siliconflatirons.org/events/2025-...

E-A-G-L-E-S

I created a new AI seminar class, and I thought I'd highlight a few readings from it each week. First week, and two great overviews of the topic: Artificial Intelligence and Law: An Overview by @harrysurden.bsky.social Paper here: scholar.law.colorado.edu/faculty-arti... and...

There has been a definite shift in recent weeks where insiders in the various AI labs are suggesting that very intelligent AIs are coming very soon. I wrote a bit about why this might be happening and what we can take away from their apparent confidence. www.oneusefulthing.org/p/prophecies...

Prophecies of the Flood

What to make of the statements of the AI labs?

oneusefulthing.org

"The key is learning to harness AI’s impressive capabilities while being clear-eyed about its limitations and the persistent necessity of human judgment in interpreting our foundational legal document," write Andrew Coan and @harrysurden.bsky.social.

AI and Constitutional Interpretation: The Law of Conservation of Judgment

Large language models cannot eliminate the need for human judgement in constitutional law.

lawfaremedia.org

Interesting interview by @sama Sam Altman with former Microsoft research head Craig Mundie. Mundie claims that a fusion startup they're invested in will produce useful fusion energy next year. I am skeptical, but who knows. Would be significant if true. www.youtube.com/watch?v=Z246...

Craig Mundie with Sam Altman | Artificial Intelligence, Hope, and the Human Spirit

YouTube video by Commonwealth Club World Affairs (CCWA)

youtube.com

@Google has released Gemini 2.0 Flash today. So far, I am not impressed in my testing suite. Its not as good as GPT-4o or Claude. However, "Flash" is the less capable turbo version of Gemini 2.0, and I expect the full version of Gemini 2 to be quite good when it is released.

Introducing Gemini 2.0: our new AI model for the agentic era

Today, we’re announcing Gemini 2.0, our most capable AI model yet.

blog.google

@OpenAI released its SORA video generation model today and it's quite impressive sora.com/explore/feat...

Interesting conversation with @OpenAI's Noam Brown @polynoamial on the new o1 AI model and the current state of AI and test-time/inference compute. www.youtube.com/watch?v=OoL8...

OpenAI's Noam Brown Unpacks the Full Release of o1 and the Path to AGI

YouTube video by Unsupervised Learning: Redpoint's AI Podcast

youtube.com

I am quite impressed with @Google's Gemini Experimental 12-06 model, released today. This is the first version of Gemini that I consider fully on par with (or perhaps exceeding) GPT-4o and Claude Claude 3.5 Sonnet, in my private AI capability test suite. cc:JeffDean

Coan & Surden on AI and Constitutional Interpretation, lsolum.typepad.com/legaltheory/...

Excellent lecture by Prof. @JustinWolfers on Artificial Intelligence and the Economy www.youtube.com/watch?v=jIL5...

AI Will Cause Jobs to Evolve, Will You Be Able to Evolve With It?│Justin Wolfers

YouTube video by World Knowledge Forum

youtube.com

I'm recruiting 1-2 PhD students to work with me at the University of Colorado Boulder! Looking for creative students with interests in #NLP and #CulturalAnalytics. Boulder is a lovely college town 30 minutes from Denver and 1 hour from Rocky Mountain National Park 😎 Apply by December 15th!

Excellent Lecture by @UCBerkeley Prof. Gabe Zucman on how the extremely wealthy avoid paying their fair share of taxes compared to typical tax payers, and how to fix it www.youtube.com/watch?v=NM61...

A New Tax System for a Globalized World│Gabriel Zucman (UC Berkeley, Associate Professor)

YouTube video by World Knowledge Forum

youtube.com

How do LLMs learn to reason from data? Are they ~retrieving the answers from parametric knowledge🦜? In our new preprint, we look at the pretraining data and find evidence against this: Procedural knowledge in pretraining drives LLM reasoning ⚙️🔢 🧵⬇️

Hot off the presses - Just posted a new draft article on "Artificial Intelligence and Constitutional Interpretation" Written with the incredible Prof. Andrew Coan papers.ssrn.com/sol3/papers.... Comments and suggestions are welcome