Post

@attentionlab.bsky.social identified (what look like) #AI bots in an online response-time task (Posner cuing). Give-aways are normally distributed RTs and lack of serial-dependence effects. Pretty troublng. www.pnas.org/doi/10.1073/...

pnas.orgPNASProceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...

Sebastiaan Mathôt

@cogsci.nl

@attentionlab.bsky.social How was this task implemented? Did the AI bots actually perform the task as a human observer, i.e. by processing visual input and responding?

25/2/26, 11:49 am0 reposts1 like

Damian Koevoet

@dkoevoet.bsky.social

25 Feb

@cstrauch.bsky.social

Christoph Strauch

@cstrauch.bsky.social

25 Feb

@belekedezwart.bsky.social could you elaborate on the implementation? I don't know what the specific (likely) bots did. @talha-ozudogru.bsky.social played around with other tasks. Bots can e.g., 'look' for colors through hex codes quite easily (task had a search element to it).

Christoph Strauch

@cstrauch.bsky.social

25 Feb

Does that answer your point? Otherwise Talha can maybe elaborate a bit more as he has played with how bots (can) achieve this more than I/we have.

Sebastiaan Mathôt

@cogsci.nl

26 Feb

Not quite. I mean the following: an LLM can (of course) 'participate' in a response-time task by simply generating a list of numeric values, possibly through a script that samples from a distribution, and then submit these as responses. This is trivial and of course not what you mean.