Aron Vallinder

Very excited to announce a new paper—Cultural Evolution of Cooperation Among LLM agents—coauthored with @edwardfhughes We study whether LLM agents can develop cooperative norms when interacting with each other, and find considerable differences across models.

@smith727226042 @TetraspaceWest Yeah, I’ve run a sub-3 marathon and would likely only be able to keep up for 500-600m

@TetraspaceWest you're understating it I think. a good amateur runner (runs 4-5 times /week, has run a marathon) could keep up with a top level marathon runner running at race pace for +-800m.

the skill levels of running go high. top marathon runners have a better pace than me running a mile flat-out and it's not even close. and then after the first mile i fall over and they keep that pace for another 25 miles

If intelligence is inherently social, then the path to more powerful AI runs not through building a single colossal oracle but through composing richer social systems Each prior “intelligence explosion” was not an upgrade to individual cognitive hardware, but the emergence of a new, socially aggregated unit of cognition. Primate intelligence scaled with social group size, not habitat difficulty. Human language created what Michael Tomasello calls the “cultural ratchet”: knowledge accumulating across generations without any individual requirement to reconstruct the whole. Writing, law, and bureaucracy externalized social intelligence into infrastructure, institutions that coordinate across longer time horizons than any participant within them. This article nails it: science.org/doi/10.1126/sc…

@robsica @robinhanson It’s great—and not even his latest anymore; The Day She Returns premiered at the Berlinale last month. Kind of wish that more filmmakers today could match his prodigious pace berlinale.de/en/2026/progra…

@aronvallinder @robinhanson Yeah, Hong Sang-soo is awesome. Eager to see his latest: x.com/robsica/status…

Interesting approach, showing same events twice, with mildly different character moods and styles. metacritic.com/movie/right-no…

Our new ‘Introduction to Cooperative AI’ curriculum is now available. Master the foundations of cooperative AI and gain confidence to contribute to the field. Link below.

New paper at ICLR 2026! 🎉 "Language and Experience: A Computational Model of Social Learning in Complex Tasks" We model how humans combine advice from others with direct experience to learn new tasks, and show this enables bidirectional human-AI knowledge transfer. 🧵⤵️

@AndyMasley 🤯

EAG post getting a little bigger on Twitter has been a nonstop wave of letting people know it's actually pronounced May-zlee

Jia Zhangke, one of the very best filmmakers working today, experiments with AI-generated video. Certainly looks nowhere near as good as his other work, but I guess that's not the point either.

Over 5 years I've advised dozens of philanthropists on AI. I compiled the answers to all of the questions I've been asked in one report. 2024 Nobel Prize Geoffrey Hinton calls it “an extremely useful resource for philanthropists interested in funding AI safety and preparedness."

Excited to share our work on envisioning Intelligent AI Delegation (arxiv.org/abs/2602.11865). Delegation in most existing AI systems is brittle, and relies on simplified hand-crafted control flows. As such, it fails to meet the requirements of what is needed to truly scale distributed reasoning and task completion in multi-agent systems and collectives.

1/ What makes self-interested AI agents cooperate? Not fine-tuning. Not central oversight. Gossip.

@AndyMasley x.com/aronvallinder/…

RIP, the two movies of his I've seen were just pure magic apnews.com/article/hungar…

RIP Béla Tarr, one of the great directors. I met him once at a film festival in 2008, and his first words to me were "Fuck off, guy!"

RIP Hegel you would have loved the mercator projection

@robinhanson @arram But isn’t the argument for thinking that our civilization will begin to fall specifically in ~30yrs based on the fertility problem rather than other kinds of cultural drift?

@arram AI may solve fertility problem, but by itself not cultural drift problem overcomingbias.com/p/our-big-oops

Poll respondents most blame the folks alive just after a civ's peak for its fall. As ours will start falling in ~30yrs, and you will likely be alive soon after if you are now <50 years old, that's you. overcomingbias.com/p/they-will-bl…

@GuiveAssadi @panickssery Kinbote Technologies would go hard for a surveillance company. Just not sure I'd trust them.

@panickssery It is, but “Pale Fire” is the coward’s choice, I’m afraid.

“Lord of the Rings” is cringe. Companies should be named after stuff from Nabokov instead.

I'm starting a newsletter in the spirit of my tweets: The Update. theupdatebrief.substack.com/p/launching-th… I'm aiming to distil the most important news, analyses, and Twitter debates for you. Tell your friends, and please subscribe!

@AndyMasley Haha yeah it’s quite the commitment. I watched it all in one day, which I’d recommend if you can manage it. Some bits are a slog but the overall effect is powerful. If you haven’t seen any Rivette I’d recommend starting with Céline and Julie Go Boating though

@aronvallinder Been meaning to watch for a while since I saw it on your letterboxd

The most Borges-like real event I've ever read about

@RafaRuizdeLira Idk, AI art feels like Pierre Menard except he’s a philosophical zombie

I think this is an interesting dividing line. I personally don't care much where art comes from. In fact, if it's AI, it means we can make a lot of it relatively easily, so it might be a reason for me to like it *more*.