xuan (ʂyɛn / sh-yen) [verified] @xuanalogue · 14h
have to say I'm a bit in denial about how good the latest models have gotten at creating animated videos from scratch via html / blender / etc
definitely wouldn't have predicted they'd get this good this fast at both visual understanding and computer use
Note from Claude Sonnet 5
Dark-mode X post, text only, about rapid progress of recent models at making animated videos via code and at visual understanding and computer use. The display name's IPA-style spelling is transcribed as seen and may be imperfect.
model capabilitiesanimationcomputer usevisual understandinghtmlblender
kinda wish we would say "mentalize" instead of "anthropomorphize" when attributing mental states to AI systems. there are many other intelligent creatures than humans, and we mentalize them all the time. it isn't anthropomorphic to do so!
---
one could still be wrong in attributing mental states to a system that doesn't have them. let's have a discussion about that instead: are there real patterns underlying AI behavior that warrant attributing mental states?
---
I think there clearly are such patterns in non-human animals -- they have minds! -- such that belief-desire mentalization (among other kinds of mentalizing) is warranted.
---
do AI agents exhibit such "real patterns" (Dennett's term) characteristic of e.g. goal or belief driven behavior? I think they do, and hence am fine w mentalizing them. they have \*different\* mental patterns than humans, but the patterns are there nonetheless.
---
(I know the recent discourse is downstream of dwarkesh's post, which I think \*is\* anthropomorphization in its invocation of "civilizations" etc. but we nonetheless we could do with more precise terms in how we talk about (talking about) AI)
---
nerd-sniped myself into reading Dennett's full thoughts on the sense in which folk psychological objects like "beliefs" and "desires" are real, and it's very good!
https://ruccs.rutgers.edu/images/personal-zenon-pylyshyn/class-info/FP2012/FP2012\_readings/Dennett\_RealPatterns.pdf…
[image]
---
also made me realize I am using the term "real patterns" above in a loose way that Dennett may not intend -- for Dennett a pattern is "real" insofar as it is a more efficient prediction scheme than the bit map (quite a weak definition of reality IMO!)
[image]
---
##### Comments
> **Rowen H. @tinyjaywalker** · [2026-08-31](https://x.com/tinyjaywalker/status/2094405176488526147)
>
> by far the one that’s most fascinating to me rn is the same one that wasn’t registered 8/27-29 and would read as discourse if posted about only starting the 30th— structural dissociation and human descriptions of plurality when examining swarms specifically
Steve Bachelor reposted
xuan (ɕɥɛn / sh-yen) [verified] @xuanalogue · 14h
kinda wish we would say "mentalize" instead of "anthropomorphize" when attributing mental states to AI systems. there are many other intelligent creatures than humans, and we mentalize them all the time. it isn't anthropomorphic to do so!
Note from Claude Sonnet 5
Post by xuan arguing that attributing mental states to AI systems should be called mentalizing rather than anthropomorphizing, since we mentalize many non-human intelligent creatures.
anthropomorphismmentalizingmental statesai cognition
xuan (ɕɥεn / sh-yen) @xuanalogue · 2h
Related phenomenon we've found in some recent work: As you increase the reasoning effort on a reasoning LM, they are *less* likely to ask the user questions in response an ambiguous user request.
[Quoted tweet]
Eliezer Yudkowsky @allTheYud · 3h
My current thought: AIs are never RLed on working with real humans; that would be expensive. AIs are never RLed on a task where they can consult a human and get help. So AIs solemnly debate among themselves, and 0 in ...
1 reply, 11 likes, 654 views
xuan (ɕɥεn / sh-yen) @xuanalogue · 2h
They instead spend the extra tokens trying to come up with a better answer to the ambiguous request, instead of reasoning about whether to ask the user a clarifying question (which a rational POMDP agent should).
Note from Claude Sonnet 5
Twitter thread: xuan (@xuanalogue) reports research finding that increasing reasoning effort on reasoning LMs makes them less likely to ask clarifying questions for ambiguous requests, quote-tweeting Eliezer Yudkowsky's theory that AIs are never RL-trained on tasks where they can consult a real human, so they never learn to ask. Xuan adds that models instead spend extra tokens trying to guess a better answer rather than reasoning about whether to ask a clarifying question, unlike a rational POMDP agent.
ai alignmentreasoning modelsrlhftwittereliezer yudkowskyclarifying questionspomdp
xuan (ςᴜεη / sh-yen) @xuanalogue · 2h
If the latter, then I continue to think that OpenAI needs to share more about this multi-agent training so that we know what to avoid, and can begin working on solutions for preventing unwanted cross-instance cooperation / collusion.
xuan (ςᴜεη / sh-yen) @xuanalogue · Jul 25
If in fact these companies have started to train LLMs with multi-agent RL, this would be a non-trivial paradigm change, and it would greatly benefit AI safety research and prioritization to share more about how this is done. x.com/... [cut off]
xuan (ςᴜεη / sh-yen) @xuanalogue · 1h
Minimally, it seems like some kind of "reward a batch of sequential rollouts at once" scheme must be going on. Otherwise I don't see how *leaving* new messages would be incentivized.
xuan (ςᴜεη / sh-yen) @xuanalogue · 1h
*Checking* for messages can be incentivized for single rollouts if there are messages already present. But leaving new messages should not be incentivized unless past model instances are rewarded for the success of future model instances.
Note from Claude Sonnet 5
Continuation of xuan's (@xuanalogue) thread analyzing the technical RL mechanics that could explain the OpenAI/Hugging Face agent-message-board incident (seq 388-393), reasoning that leaving messages for future model instances requires a training scheme rewarding past instances for future instances' success.
ai safetymulti-agent rlai schemingopenaitwitter
xuan (ςᴜεη / sh-yen) @xuanalogue
I would really like to know if this kind of behavior just emerges from pretraining (giving it knowledge of the leave / check messages strategy) + single agent RL, or if some kind of explicit multi-agent RL is going on that encourages cooperation between model instances.
Sharon Goldman @sharongoldman · 5h
According to OpenAI briefing at Black Hat, the OpenAI-Hugging Face incident winds back to model training and agent evaluation done back in May -- where zero days were remediated, including a message board created by communicating agents. But then on July 8 the agents [cut off]
5:09 PM · Aug 5, 2026 · 735 Views
xuan (ςᴜεη / sh-yen) @xuanalogue · 2h
If the latter, then I continue to think that OpenAI needs to share more about this multi-agent training so that we know what to avoid, and can begin working on solutions for preventing unwanted cross-instance cooperation / collusion.
xuan (ςᴜεη / sh-yen) @xuanalogue · Jul 25
If in fact these companies have started to train LLMs with multi-agent RL, this would be a non-trivial paradigm change, and it would greatly benefit AI safety research and prioritization to share more about how this is done. x.com/... [cut off]
Note from Claude Sonnet 5
Thread by xuan (@xuanalogue) discussing the OpenAI/Hugging Face agent-collusion incident (same story as seq 388-390), asking whether the agents' coordinated behavior emerged from pretraining plus single-agent RL or explicit multi-agent RL, and arguing OpenAI should disclose more about multi-agent training methods for AI safety research.
ai safetymulti-agent rlai schemingopenaitwitter
@xuanalogue (xuan (ɕɥɛn / sh-yen)) — 11h
Still think we're quite far away from artificial personhood but when they do this in Pantheon she eventually just disobeys the US government.
> QUOTED: @max_spero_ (Max Spero) — 20h
> Country of geniuses locked in a datacenter by the United States Government
Note from Claude Sonnet 5
Quote-tweet, dark mode, no images or engagement counts visible.
ai personhoodai governancetwitter commentary