professor at university of washington and scientist at Google DeepMind. computational cognitive scientist. working on social and artificial intelligence and alignment. http://faculty.washington.edu/maxkw/https://bsky.app/profile/maxkw.bsky.socialhttps://bsky.app/profile/maxkw.bsky.social/post/3mypjmeibm225Excited by our new work on AI contracting! Now that we each have our own agents, they need to be able to find mutual benefit with each other and honor their agreements. This is a promising route to supercooperators! [contains quote post or other embedded content]25 Oct 2026 15:30 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mypjmeibm225https://bsky.app/profile/maxkw.bsky.social/post/3mxifywijj22jWe won best paper at the COLM Workshop on Agent Behavior for our paper: AI Assistants Overassist! Congratulations @veronateo.bsky.social, Raghav Jain and @tobigerstenberg.bsky.social! Check out the preprint arxiv.org/abs/2607.2130610 Oct 2026 02:12 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mxifywijj22jhttps://bsky.app/profile/maxkw.bsky.social/post/3mwvr3rf7ks2rEveryone is excited about recursive *self* improvement (RSI) but human progress is social and driven by cumulative culture. Individuals stand on the shoulders of past giants. Our new paper tests whether LLM swarms benefit from an alternative RSI: recursive *social* improvement. [contains quote post or other embedded content]02 Oct 2026 16:09 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mwvr3rf7ks2rhttps://bsky.app/profile/maxkw.bsky.social/post/3mvaskes6cc2jOur newest work unifying game theoretic ideas about the evolution of cooperation with the emergence of computation itself. Self-replicating programs first emerge through random mutation and then spread by cooperating. [contains quote post or other embedded content]11 Sep 2026 14:44 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mvaskes6cc2jhttps://bsky.app/profile/maxkw.bsky.social/post/3msgv2fthcc2pNew lab preprint! AI assistants are myopically helpful. Unlike caregivers that want to empower people and grow their long term autonomy, AI assistants interrupt independent thinking and jump right to the answer. [contains quote post or other embedded content]06 Aug 2026 20:30 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3msgv2fthcc2phttps://bsky.app/profile/maxkw.bsky.social/post/3mr7bociqh226Just arrived in Rio for #CogSci2026! I'll be at the Cognitive Science of AI Alignment workshop on Wednesay afternoon to talk about "Machines That Care Like Us"22 Jul 2026 02:29 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mr7bociqh226https://bsky.app/profile/maxkw.bsky.social/post/3mnkpufbpzs26Excited about our new work measuring multi-turn persuasion in AI-human interactions and how to simulate human persuadability! [contains quote post or other embedded content]05 Jun 2026 18:01 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mnkpufbpzs26https://bsky.app/profile/maxkw.bsky.social/post/3mg4fwhorss2qOur new short piece in TiCS on intuitive theories of truth: how people judge whether statements could be true, whether statements are true, and whether to assert them as true. A great collab with @keremoktar.bsky.social @ihandleyminer.bsky.social @kevinzollman.com @lianeleeyoung.bsky.social [contains quote post or other embedded content]02 Mar 2026 23:30 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3mg4fwhorss2qhttps://bsky.app/profile/maxkw.bsky.social/post/3m2bbyn5w7k2qNew paper challenges how we think about Theory of Mind. What if we model others as executing simple behavioral scripts rather than reasoning about complex mental states? Our algorithm, ROTE (Representing Others' Trajectories as Executables), treats behavior prediction as program synthesis. [contains quote post or other embedded content]03 Oct 2025 05:01 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3m2bbyn5w7k2qhttps://bsky.app/profile/maxkw.bsky.social/post/3m2a757uwc22lWhen values collide, what do LLMs choose? In our new paper, "Generative Value Conflicts Reveal LLM Priorities," we generate scenarios where values are traded off against each other. We find models prioritize "protective" values in multiple-choice, but shift toward "personal" values when interacting. [contains quote post or other embedded content]02 Oct 2025 18:37 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3m2a757uwc22lhttps://bsky.app/profile/maxkw.bsky.social/post/3m2476do3bs2eExcited by our new work estimating the empowerment of LLM-based agents in text and code. Empowerment is the causal influence an agent has over its environment and measures an agent's capabilities without requiring knowledge of its goals or intentions.01 Oct 2025 04:27 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3m2476do3bs2ehttps://bsky.app/profile/maxkw.bsky.social/post/3lvrceynjp227Claire's new work showing that when an assistant aims to optimize another's empowerment, it can lead to others being disempowered (both as a side effect and as an intentional outcome)! [contains quote post or other embedded content]06 Aug 2025 22:44 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3lvrceynjp227https://bsky.app/profile/maxkw.bsky.social/post/3lujthpas3c22Our new paper is out in PNAS: "Evolving general cooperation with a Bayesian theory of mind"! Humans are the ultimate cooperators. We coordinate on a scale and scope no other species (nor AI) can match. What makes this possible? 🧵 https://www.pnas.org/doi/10.1073/pnas.240099312222 Jul 2025 06:03 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3lujthpas3c22https://bsky.app/profile/maxkw.bsky.social/post/3lu54mffams27Very excited for this workshop! [contains quote post or other embedded content]17 Jul 2025 04:42 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3lu54mffams27https://bsky.app/profile/maxkw.bsky.social/post/3lsrr47idj22vSettling in for my flight and apparently A.I. DOOM is now a movie genre between Harry Potter and Classics. Nothing better than an existential crisis with pretzels and a ginger ale.29 Jun 2025 22:52 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3lsrr47idj22vhttps://bsky.app/profile/maxkw.bsky.social/post/3lr77hsnd7k2zLLMs learn beliefs and values from human data, influence our opinions, and then reabsorb those influenced beliefs, feeding them back to users again and again. We call this the "Lock-In Hypothesis" and develop theory, simulations, and empirics to test it in our latest ICML paper!09 Jun 2025 20:23 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3lr77hsnd7k2zhttps://bsky.app/profile/maxkw.bsky.social/post/3lnvopuji7c2mExcited to speak about some new work on Bayesian Cooperation at this workshop! Join us virtually [contains quote post or other embedded content]28 Apr 2025 21:14 +0000at://did:plc:jbopvttj6tmoradsay7qkxj3/app.bsky.feed.post/3lnvopuji7c2m