every few years i say “what if we could automate psychology research?” this time i think we’ve made real progress (because ai).
New preprint! AI agents have shown impressive scientific automation capabilities. Can we apply them to psychology research, *including* human data collection?
Introducing auto-psych, a framework that proposes cognitive models and uses them to design and run human experiments. 1/
Users are excited about the Auto-Psych framework automating psychology experiments because of its fun origin story and the talented team of researchers and students who built it.
No Digg Deeper questions have been answered for this story yet.
i just posted Stanford auto-psych project. here’s simultaneous awesomeness from our friends at Princeton!
1/ 🚨 New preprint: "Closing the Loop to Discover Psychological Theories with an Automated Cognitive Scientist"
Co-led w/ Younes Strittmatter Co-mentored by @suyoghc and @cocosci_lab In collaboration w/ @kachergis, @norijacoby, @nathanieldaw, & @cpilab
Introducing AutoCog — a fully autonomous AI system that runs the entire scientific discovery cycle in cognitive science to surface novel theories of human behavior 🧵
#CognitiveScience #AI4Science #LLMs
If you don't know what a nerd swipe is, well: http://xkcd.com/356/ - so I couldn't stop thinking about his suggestion. On a long car ride by myself I had a very long think about the topic and began planning what became auto-psych.
fun origin story: i nerd-sniped @mcxfrank a few months ago by saying “i think this is feasible now”. he came back with a beautiful agentive flow for auto-psych. this got me curious how much could be just replaced by coding agents… and eventually a crew jumped in to awesomify
every few years i say “what if we could automate psychology research?” this time i think we’ve made real progress (because ai).
Fun story about this paper, which I'm really excited about! This came out of a classic nerdswipe by @noahdgoodman - we were teaching together this winter and he said "is it time for robots to do psychology?"
New preprint! AI agents have shown impressive scientific automation capabilities. Can we apply them to psychology research, *including* human data collection?
Introducing auto-psych, a framework that proposes cognitive models and uses them to design and run human experiments. 1/
I spent a frantic couple of weeks vibe coding a draft MVP then ran out of time, but I posted enough sufficiently funny things about it that Noah got nerd-swiped in turn and started porting my very rigid pipeline into a more flexible, agent-based workflow.
Then we were very lucky that a super talented group of students, led by Ben Prystawski, stepped in and formalized the pipleine, making it a real Bayesian theory discovery loop and launching what we think are the first agent-driven psychology experiments!
@mcxfrank haha..love it! same goes for us. Younes,@suyoghc, and I thought we can coded it up in a weekend but took a few more than we thought :)
TIL about nerd sniping! Yep, similar story here! The nerd-sniping feedback loop and contagion started a bit earlier with @akjagadish and @kachergis resulting in related work (https://arxiv.org/abs/2604.25521), but also spread quickly to Younes & many co-authors who'd been thinking about these topics for a while!
Main symptoms imo: obsession and loss of sleep from trying "just one more thing", expanding the scope and rabbit-holing to adjacent questions, then trying to narrow it down, and finally having "we're almost done" become a somewhat chronic state!
@mcxfrank @noahdgoodman Neat!
@noahdgoodman Cheers for the (joint) awesomeness to follow! 🍻