Lisa Messeri
banner
lmesseri.bsky.social
Lisa Messeri
@lmesseri.bsky.social
anthropologist of sci & tech. Prof @Yale. author of "Placing Outer Space" and VR book "In the Land of the Unreal". tech criticism with good vibes.
Pinned
Can the human sciences exist w/out the human? Proposals for using AI as human research subjects suggest yes. But @mjcrockett.bsky.social and I respond with, ‘not so fast.’ In fact, silicon subjects say more about the problems of the research paradigm than the promises of AI. 🚨New article, thread 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
Reposted by Lisa Messeri
This NeurIPS workshop claims that LLMs "provide an important foundation for exploring human cognition, emotion, and social interaction"

This is flawed logic, as @lmesseri.bsky.social and I argue here:
www.sciencedirect.com/science/arti...
November 9, 2025 at 3:23 PM
Reposted by Lisa Messeri
Excited to welcome @lmesseri.bsky.social, @cameronbuckner.bsky.social and @carlbergstrom.com to campus a week from today for the “AI and the Nature of Science: Concepts and Controversies” event. 14.11.25, 1400 - 1800.
AI and the Nature of Science
Bringing together cutting-edge perspectives and helping set an agenda for future research and application
buff.ly
November 7, 2025 at 2:45 PM
Reposted by Lisa Messeri
Pop-up AI defacing your article critical of AI captures a lot of what it feels like to work in this space.
@lmesseri.bsky.social
November 7, 2025 at 3:47 PM
Reposted by Lisa Messeri
Important work revealing limitations of using LLMs as human surrogates.

Note, however, that even if LLMs' textual outputs were perfectly human-like, they would still be poor models of human cognition, as @lmesseri.bsky.social argue here:
www.sciencedirect.com/science/arti...
November 7, 2025 at 3:39 PM
Yesterday's NYTimes platforming of the heinous perspective that women have ruined the workforce was bad. But to post a photo of the the winner of the Great British Bake Off first thing Friday morning? Have you no respect for your readers?
November 7, 2025 at 12:55 PM
Yale Anthropology is doing an open rank search for environmental anthropology. Happy to answer questions in DMs (I am not on the search committee). We’re a lovely department (imo) and given the current higher ed atmosphere, Yale has been doing a decent job at getting us thru (knock on wood).
Assistant/Associate/Full Professor, Environmental Anthropology - New Haven, Connecticut job with Yale University, Department of Anthropology | 704970
The Department of Anthropology at Yale University seeks to hire a full-time tenure-track or tenured assistant/associate/full professor in environme...
careercenter.americananthro.org
November 6, 2025 at 3:07 PM
I am participating and reviewing submissions for this workshop. so let's get some interesting proposals here, yeah? If you are an STSer/anthropologist working on how AI is changing ideas and practices in the sciences, would love to have you in the conversation.
The Craft of Science with AI: Evidence, Judgment, and Practice
We invite researchers and practitioners to join us in examining how scientific reasoning and imagination are being reconfigured as AI systems become a part of the everyday practice of science. Learn m...
datasociety.net
November 6, 2025 at 2:44 PM
I don't write about outer space much these days, but I do have two new essays out on the topic. The first is a short afterward for a recent collection of critical outer space studies. I wonder why our space imagination has become so narrow and draw inspiration from NK Jemisin's Broken Earth trilogy
Afterword: Critique and Imagination: Re-Connecting with the Infinit...
On 29 April 2025, the James Web Space Telescope released an image titled “A Visual Feast of Galaxies.” The image is densely packed with pinpricks of light, most of which, upon closer inspection, tr...
journals.openedition.org
November 6, 2025 at 2:25 PM
Reposted by Lisa Messeri
Yay! Out now!! Thanks to @lmesseri.bsky.social @niiyokamigaabaw.bsky.social, Alexandra Ganser, Stefanie Dunning, Natalie Treviño, Matt Harvey and Frédéric Boone for their wonderful contributions and afterwords!
November 5, 2025 at 3:13 PM
For the Minnesotans out there, see post below for upcoming book talk! My host, @justcode.bsky.social, wrote a lovely review of my book right when it came out. Also it's the final days of the @dukepress.bsky.social Fall book sale (thru Nov 9). Use FALL25 code on the press website to get 50% off!
November 4, 2025 at 6:20 PM
Some Sunday reading, if you are so inclined (open access pre-print linked in the thread comments
👇
Can the human sciences exist w/out the human? Proposals for using AI as human research subjects suggest yes. But @mjcrockett.bsky.social and I respond with, ‘not so fast.’ In fact, silicon subjects say more about the problems of the research paradigm than the promises of AI. 🚨New article, thread 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
November 2, 2025 at 6:40 PM
Reposted by Lisa Messeri
Love it, seasonally appropriate, no notes
We draw a parallel to WEIRD, proposing that CogSci is DEAD (Decontextualized, Engineered, Anonymized, and Disembodied). This is meant to be provocative, of course, but, like WEIRD, it is in the spirit of illuminating limitations and striving for better science 6/
October 25, 2025 at 12:27 AM
stubbled across a short NYT piece by Toni Morrison from 1973. This line stood out: "For this year fantasy itself lost its genuineness. The normal lines of communication between sham and reality had broken down."
On to Disneyland and the Real Unreality (Published 1973)
www.nytimes.com
October 24, 2025 at 12:02 PM
Reposted by Lisa Messeri
Spotted on LinkedIn... a bad AI summary of our new paper on risks of AI in research.

Please make it stop.
October 23, 2025 at 9:58 PM
Reposted by Lisa Messeri
'New research coordinated by the European Broadcasting Union (EBU) and led by the BBC has found that AI assistants – already a daily information gateway for millions of people – routinely misrepresent news content no matter which language, territory, or AI platform is tested.' 1/2
Largest study of its kind shows AI assistants misrepresent news content 45% of the time – regardless of language or territory
An intensive international study was coordinated by the European Broadcasting Union (EBU) and led by the BBC
www.bbc.co.uk
October 23, 2025 at 1:47 PM
Reposted by Lisa Messeri
I’ll say it straight: It is unbecoming, undignified, cowardly, and shameful for a institution as revered and respected as the University of Virginia to grant an audience in 2025 to those who are committed to destroying science, free thought, human rights, and the rule of law.
October 22, 2025 at 6:20 PM
Reposted by Lisa Messeri
A new article by @lmesseri.bsky.social and @mjcrockett.bsky.social whose crucial work on the illusion of scientific understanding among AI zealots cannot be taught and discussed enough!
We look forward to reading this new work.
Can the human sciences exist w/out the human? Proposals for using AI as human research subjects suggest yes. But @mjcrockett.bsky.social and I respond with, ‘not so fast.’ In fact, silicon subjects say more about the problems of the research paradigm than the promises of AI. 🚨New article, thread 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
October 21, 2025 at 10:19 PM
Reposted by Lisa Messeri
Another brave article by @lmesseri.bsky.social and @mjcrockett.bsky.social . Essential reading if you care about the harm AI celebrationism is doing!
Can the human sciences exist w/out the human? Proposals for using AI as human research subjects suggest yes. But @mjcrockett.bsky.social and I respond with, ‘not so fast.’ In fact, silicon subjects say more about the problems of the research paradigm than the promises of AI. 🚨New article, thread 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
October 22, 2025 at 11:50 AM
Reposted by Lisa Messeri
This study shouldn't have been necessary. In fucking cognitive science, no less. I'd say I couldn't believe it, but @avastmachine.bsky.social showed the historical connections in his book The Closed World.

But since it was necessary, I'm very glad it happened.

Will cog sci pay sufficient attn?
Can AI simulations of human research participants advance cognitive science? In @cp-trendscognsci.bsky.social, @lmesseri.bsky.social & I analyze this vision. We show how “AI Surrogates” entrench practices that limit the generalizability of cognitive science while aspiring to do the opposite. 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
October 21, 2025 at 9:24 PM
This is the way (for me). Piloted it with my graduate seminar this semester. I've been marking reading responses by hand, returning them the following session. Unlike what i had been doing for my whole career, which was feeling like i had to mark immediately and return through Canvas.
Fellow academics, I'm doing it. Going all in. Pulling the plug, literally.

This is on my syllabus for Spring 2026: "This is an analog class. No Canvas; this syllabus is your guide."

All readings are printed books; hand-written exams; no laptops. #academicbluesky #Highered #academia #Rutgers
October 21, 2025 at 9:47 PM
OpenAI: Hallucinations are mathematically inevitable.

Also OpenAI: here's an AI web browser.

www.nytimes.com/2025/10/21/t...
OpenAI Unveils Web Browser Built for Artificial Intelligence
www.nytimes.com
October 21, 2025 at 9:26 PM
Reposted by Lisa Messeri
Can AI simulations of human research participants advance cognitive science? In @cp-trendscognsci.bsky.social, @lmesseri.bsky.social & I analyze this vision. We show how “AI Surrogates” entrench practices that limit the generalizability of cognitive science while aspiring to do the opposite. 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
October 21, 2025 at 8:24 PM
Reposted by Lisa Messeri
Huge thanks to @mjcrockett.bsky.social and @lmesseri.bsky.social for digging into all that work to produce this survey and their insights. The problems that cogsci research is facing are real, but "AI surrogates" are not a solution.
Can AI simulations of human research participants advance cognitive science? In @cp-trendscognsci.bsky.social, @lmesseri.bsky.social & I analyze this vision. We show how “AI Surrogates” entrench practices that limit the generalizability of cognitive science while aspiring to do the opposite. 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
October 21, 2025 at 8:35 PM
This link to the article (published in Trends in Cognitive Science, I forgot to mention) should be OA for the next 50 days: authors.elsevier.com/c/1lzmh4sIRv...
October 21, 2025 at 8:11 PM
Reposted by Lisa Messeri
in which @lmesseri.bsky.social and @mjcrockett.bsky.social upend the increasingly pervasive premise of the LLM-fabricated human research subject (and the histories/ ideologies of cognition baked into the experimental "tasks" they're called upon to complete) at its epistemic roots
Can the human sciences exist w/out the human? Proposals for using AI as human research subjects suggest yes. But @mjcrockett.bsky.social and I respond with, ‘not so fast.’ In fact, silicon subjects say more about the problems of the research paradigm than the promises of AI. 🚨New article, thread 1/
AI Surrogates and illusions of generalizability in cognitive science
Recent advances in artificial intelligence (AI) have generated enthusiasm for using AI simulations of human research participants to generate new know…
www.sciencedirect.com
October 21, 2025 at 4:43 PM