Sign in

Rafael M Batista

@rafmbatista.bsky.social
4.5K followers 1.3K following 530 posts

Behavioral Scientist. Exploring how AI is shaping the way we experience the world. New DC resident + civically engaged, so occasionally I post about that too.

PostsRepliesMedia
Rafael M Batista @rafmbatista.bsky.social · 2h
I'm probably not the right audience for the benchmark papers because I really don't understand them. If I saw psychologists popping out scales out as quickly as LLM benchmarks, I'd be even more skeptical than I am of them today. What's the idea here?
010
Rafael M Batista @rafmbatista.bsky.social · 9h
I spoke with @npr.org this week about #AI sycophancy and some things to be aware of if you turn to LLMs this election season. You can check it out here: www.npr.org/2026/10/04/n...
npr.org
Midterm voting is underway. Some voters are relying on AI to make a decision
Voters are already voting in the midterms. This year, some voters are trying something new to get ready for the election: asking AI to help research their ballot and even decide who to vote for.
192
Reposted by Rafael M Batista
Cole Donovan @colesci.bsky.social · 11h
EXTREMELY IMPORTANT DATA CALL ALERT: The OECD is running a survey on researcher career intentions among the member economies. Right now, I'm told that the response rate on the US side is low (to the point that US data may not get published). We need inputs from reliable sources at this moment.
survey.oecd.org
OECD-EU Research and Development Careers Survey 2026
1613
Reposted by Rafael M Batista
Maria Antoniak @mariaa.bsky.social · 03/10/2026
Maybe worth saying aloud that I do worry about safety risks (hacking, bio, an “out of control” agent). I just also see those risks as deeply entangled with money, politics, personalities, regulation, such that I doubt we can “solve safety” if we don’t first address foundational issues.
3678
Reposted by Rafael M Batista
Alondra Nelson @alondra.bsky.social · 03/10/2026
In the Biden administration, l led the team that developed what was arguably the first White House “AI safety” framework, the AI Bill of Rights. David Robinson, author of this forthright and sobering essay about risk culture at OpenAI and in Silicon Valley was on that team.
10407179
Rafael M Batista @rafmbatista.bsky.social · 03/10/2026
Firms & govt agencies everywhere are "DOGE-ing" in the name of efficiency, but any short-term gains may come w/ long-term costs. This work draws on distributed computing + cog sci, to show how redundancy *improves* efficiency & makes orgs more resilient doi.org/10.31234/osf...
doi.org
Redundancy Protects Human Collaborations From Failure
In the name of efficiency, organizations and governments around the world are increasingly trimming redundancy---personnel or overlapping capabilities beyond what routine operations require. Drawing on distributed computing and cognitive science, we challenge this view.
000
Rafael M Batista @rafmbatista.bsky.social · 03/10/2026
Imagine you order delivery at 6:00. What feels longer, 'Your delivery will arrive in 6 min' or '...will arrive at 6:06'? What happens when it's longer, '...delivery arriving in 75 min' vs `...arriving at 7:15'? New @jcrnews.bsky.social paper by two friends Jiabi Wang & Kristin Donnelly 🥳
130
Rafael M Batista @rafmbatista.bsky.social · 03/10/2026
Love this work
131
Reposted by Rafael M Batista
Gillian Hadfield @ghadfield.bsky.social · 02/10/2026
Thank you to Prime Minister @mark-carney.bsky.social for asking me to join his new National Council on Artificial Intelligence, which will advise on Canada’s AI for All strategy. buff.ly/cNVtSiY
buff.ly
Prime Minister Carney launches new National Council on Artificial Intelligence
AI for All is the government’s new national strategy to ensure that AI is adopted responsibly, in a way that truly serves all Canadians – building trust, expanding opportunities, and reinforcing our…
231
Rafael M Batista @rafmbatista.bsky.social · 02/10/2026
New @aeon.co piece out today by my favorite writer / academic / wife @mkang.bsky.social aeon.co/essays/early...
aeon.co
Early violence is a bad teacher. But the mind can learn anew | Aeon Essays
A world of violence sustains itself through guns, poverty and other social ills, but also through the minds it creates
022
Rafael M Batista @rafmbatista.bsky.social · 02/10/2026
New paper published in PNAS finds that individuals consulting an AI with a hidden objective led to a meaningful shift in their preferences. And yet, individuals still felt the AI was helpful. One study (n=233), but important consideration for AI governance doi.org/10.1073/pnas...
doi.org
Human preferences are susceptible to covertly misaligned AI advice | PNAS
AI assistants are increasingly used as advisors to guide decisions, yet little is known about how people evaluate such advice when the advisor’s un...
141
Rafael M Batista @rafmbatista.bsky.social · 02/10/2026
In the near future—as in now—I think the bigger #AI risk is not coming from the truly remarkable models but from the same-as-always humans 🙈 OpenAI has forever changed the world. But as a firm, in *this* society, it’s still got its training wheels on www.wired.com/story/a-flaw...
wired.com
A Flaw in ChatGPT’s Mac App Could Have Let Hackers Grab Sensitive Data
While the focus has been on AI agents’ hacking capabilities, a recently patched vulnerability in a ChatGPT app shows that AI software is itself an inviting—and vulnerable—target.
220
Rafael M Batista @rafmbatista.bsky.social · 01/10/2026
Curious about a research position at leading #AI lab, I uploaded cv to pre-fill the app. Every presentation I've given at a university was logged as a separate degree 😑 and this is at a frontier AI firm why do we judge AI capabilities by their ceiling and not their floor?
001
Rafael M Batista @rafmbatista.bsky.social · 29/09/2026
I wonder how many books we’d read if all of our push notifications turned to excerpts of a novel. Can you imagine? Reading a book through these bite-sized chunks 🤓📚
020
Rafael M Batista @rafmbatista.bsky.social · 28/09/2026
I read a paper today that had a study which the author pre-registered. The results did not support the pre-registered hypotheses. This was disclosed in the paper (which is good), but then the paper “treats” it as “exploratory”. Not sure what to make of that.
110
Rafael M Batista @rafmbatista.bsky.social · 27/09/2026
Not looking good for #OpenAI. I’m guessing they will slow things down, especially if investors get spooked. Each incident disclosed is worse than the previous one. www.nytimes.com/2026/09/25/t...
nytimes.com
OpenAI’s A.I. Went Rogue and Meddled With U.S. Government Websites
The company did not learn until recently that its technology had meddled with websites for the Education and Commerce Departments and the Securities and Exchange Commission.
001
Reposted by Rafael M Batista
Gillian Hadfield @ghadfield.bsky.social · 25/09/2026
The recording of my Next Conversations panel with Stuart Russell and @deanwb.bsky.social at the Hopkins Bloomberg Center is up. We don't agree on everything, but we agree the gap between what these systems can do and the tools we have to keep them in check is widening fast.
101
Reposted by Rafael M Batista
David Mimno @dmimno.bsky.social · 22/09/2026
It's possible for Jev/Laya/Decision Models to be not that big a deal as tech and massive as a new paradigm. Here's why I'm really excited from an NLP history perspective (thread)
17619
Reposted by Rafael M Batista
Alison Gopnik @alisongopnik.bsky.social · 23/09/2026
A new podcast in the Early Childhood Matters series - my new thoughts about grandparents and kids with shout outs to Kristen Hawkes grandmother hypothesis and Michael Gurven on elders as teachers. earlychildhoodmatters.online/2026/the-gra...
earlychildhoodmatters.online
The Grandmother Hypothesis with Alison Gopnik on Birth of a Parent
Alison Gopnik explores the grandmother hypothesis and why raising children was never meant to be a job for parents alone.
091
Rafael M Batista @rafmbatista.bsky.social · 23/09/2026
For anyone currently living in Chicago or moving there in the coming months, I'm selling my apt in Hyde Park 🏠️ 5 min walk to UChicago and Obama Center / Jackson Park. Metra line next door. 1000sqft 1b/1b unit w/ plenty of natural light 🌞 www.zillow.com/homedetails/...
zillow.com
1534 E 59th St #L2, Chicago, IL 60637
Welcome to Midway Apartments and this spacious one-bedroom unit, where vintage character and gracious proportions come together beautifully. High ceilings, handsome oak floors, plaster moldings, and vintage lighting are among the home's many distinctive features. Built in 1925 and designed by noted Chicago architect Paul Frederick Olsen, Midway Apartments is a beautifully maintained courtyard co-op is distinguished by limestone and metal architectural details, with inviting landscaping. The property is ideally situated just steps from the Obama Presidential Center, the University of Chicago, Promontory Point and a wealth of neighborhood amenities.
011
Reposted by Rafael M Batista
Katy Milkman @katymilkman.bsky.social · 23/09/2026
‼️New Paper Alert‼️What does it take to change behavior? In a fresh Annual Review, Jan Voelkel, @angeladuckworth.bsky.social & I argue it's helpful to consider 3 phases of change: 1️⃣ building motivation 2️⃣ following-through 3️⃣ forming durable habits 👀 Take a Look: tinyurl.com/change-annua...
0143
Rafael M Batista @rafmbatista.bsky.social · 23/09/2026
Chicago Booth School of Business is hiring for tenure track positions in #BehSci. They start reviewing applications on September 15, 2026 apply.interfolio.com/191071
apply.interfolio.com
Apply - Interfolio
010
Rafael M Batista @rafmbatista.bsky.social · 23/09/2026
Why a gap in p(doom) between superforecasters vs. #AISafety community? I think @jackstilgoe.bsky.social has an evergreen take "...faced with extreme uncertainty and intense societal interest, the [AI Safety] experts are imagining risks in ways that fit their expertise." doi.org/10.1126/scie...
doi.org
Technological risks are not the end of the world
There’s a scene in the movie Oppenheimer in which the protagonist is trying to explain to General Groves, his military overseer, the hazards of their endeavor. Groves asks Oppenheimer, “Are you…
010
Reposted by Rafael M Batista
Jack Stilgoe @jackstilgoe.bsky.social · 23/09/2026
Interested to see (via @timharford.ft.com) the new Automated AI Risk Observatory. forecastingresearch.org/research/airo. I mentioned Tetlock's earlier efforts in this piece, noting the difference between 'superforecasters' and self-appointed existential risk people www.science.org/doi/10.1126/...
science.org
Technological risks are not the end of the world
There’s a scene in the movie Oppenheimer in which the protagonist is trying to explain to General Groves, his military overseer, the hazards of their endeavor. Groves asks Oppenheimer, “Are you saying...
041
Reposted by Rafael M Batista
Stephan Lewandowsky @lewan.uk · 20/09/2026
Can people hold incompatible conspiracy beliefs at once—for example, that Princess Diana was murdered and faked her death? A new paper authored by my team, with Alessandro Miani in the lead, finds the answer is often yes, but the key is measuring incoherence properly. doi.org/10.1093/pnas... 🧵 1/9
26330
Reposted by Rafael M Batista
Michael Clemens @mclem.org · 19/09/2026
This miracle is the fruit of generations of enormous and bipartisan investments of public research support for our universities. Future versions of this miracle are what we are currently losing, due to the US Administration’s decision to shatter the public’s partnership with our universities. 
018789
Reposted by Rafael M Batista
David Rand @dgrand.bsky.social · 16/09/2026
🚨New WP: Protecting users from AI persuasion🚨 🔸A 1-paragraph AI literacy treatment (explaining AIs can be told to pursue non-accuracy goals/to persuade) cuts AI dialogue political persuasion by ~half! 🔸No sig effect on general genAI trust arxiv.org/abs/2609.16432 w/ @rorchinik.bsky.social
24015
Reposted by Rafael M Batista
Thomas Talhelm @thomastalhelm.bsky.social · 16/09/2026
🚨 It's published! 🚨 This is the big one. 100 cultures, 12,000 people, 108 researchers. Why people (and our most-used surveys) get collectivism wrong and how to fix it. www.nature.com/articles/s41... @natureportfolio.nature.com
312462
Rafael M Batista @rafmbatista.bsky.social · 16/09/2026
I don’t understand why we’re spending this money—what’s the point of this war?
Screenshot reads “The war in Iran has cost America $38bn (or around 0.1% of GDP) so far, according to an estimate from the Congressional Budget Office. The nonpartisan scorekeeper reckons the conflict will cost America an additional $2bn-3bn each month, and projected that high energy costs will increase inflation by 0.5% in the first quarter of 2027. It said the Pentagon did not respond to its requests for information.”
000
Rafael M Batista @rafmbatista.bsky.social · 15/09/2026
I love em-dashes and I love footnotes. Both are such useful writing tools and they make for clearer writing. LLMs have ruined em-dashes for all of us. If they decide to get into the footnote game, we’re toast.
120
Reposted by Rafael M Batista
Dan Björkegren @dbjork.bsky.social · 10/08/2026
We're hiring an assistant professor in AI/data science and public policy at Columbia University! For best results apply by Sept 30 apply.interfolio.com/190715
apply.interfolio.com
Apply - Interfolio {{$ctrl.$state.data.pageTitle}} - Apply - Interfolio
032
Reposted by Rafael M Batista
Kunal Jha @kjha02.bsky.social · 11/09/2026
Can self-interested, self-improving, self-replicating agents learn to cooperate? Our new paper, Tapes Together Strong, shows they can: when social behavior, computation, and reproduction share one energy budget, cooperation evolves from scratch. arxiv.org/abs/2609.10817 🧵
48418
Reposted by Rafael M Batista
Santa Fe Institute @sfiscience.bsky.social · 14/09/2026
Applications for the 2027 Complexity Postdoctoral Fellowships are open! Marina Dubova just completed SFI’s Complexity Postdoctoral Fellowship, and she reflects on what the experience meant to her and her research. Deadline: September 30, 2026 Requirements & application: santafe.edu/sfifellowship
03115
Rafael M Batista @rafmbatista.bsky.social · 15/09/2026
What in the world… you’re telling me METR is gifted hundreds of thousands of dollars worth of API credits by some unnamed AI company that they may or may not be “independently” evaluating?
Screenshot reads: “METR said the accrued credits would have racked up approximately $600,000 in bills had it not been provided to the non-profit for free by the model provider. It did not name the AI company.”
100
Reposted by Rafael M Batista
Maria Antoniak @mariaa.bsky.social · 14/09/2026
We're hiring in Computer Science at the University of Colorado Boulder! ☀️⛰️ Machine learning and NLP people, please apply!
05531
Rafael M Batista @rafmbatista.bsky.social · 13/09/2026
I’ve got a @semble.so Collection of #AI tools I’ve come across specifically for academic research. Most of these Ive tried and found them user-friendly-ish. If you use something not on the list, please share it :) semble.so/profile/rafm...
semble.so
AI Tools for Academic Research (by Rafael M Batista) — Semble
View Rafael M Batista's collection on Semble
195
Reposted by Rafael M Batista
NBER @nber.org · 13/09/2026
Foundational expertise may be a prerequisite for extracting durable skill from AI-assisted practice, from David Autor, Tanya Rodchenko, Josh Martin, Zanna Iscenko, Scott Strand, David Pearl, and Melissa Ferere www.nber.org/papers/w35720
02616
Reposted by Rafael M Batista
Mark Riedl @markriedl.bsky.social · 12/09/2026
This is a correct take imo. First, METR and Anthropic are all part of the same rationalist/longtermist community. This is akin to giving your friends money. Second, paying METR means they have a financial interest in certain outcomes. This is not the same as government regulatory oversight
18417
Rafael M Batista @rafmbatista.bsky.social · 12/09/2026
I wonder if part of the aura of success surrounding the frontier model capabilities stem from tasks where success is defined by the output. This is great for many engineering problems where the ‘how’ doesn’t matter (it’s referred to as an “engineering mindset”).
110
Rafael M Batista @rafmbatista.bsky.social · 12/09/2026
We shouldn't expect to achieve "alignment" with #AI. There's an infinite set of community norms + values to align to, and the recent mathematics achievements promoted by OpenAI illustrate how misaligned the human engineers already are to the experts
mathandai.org
Declaration — Math and AI
Read the declaration and add your name.
100
Reposted by Rafael M Batista
NBER @nber.org · 12/09/2026
When experts know more than their evidence proves, verifiability trades credibility for flexibility: it aids communication under conflict but hinders it under aligned preferences, from Alessandro Lizzeri, Yichuan Lou, and Jacopo Perego www.nber.org/papers/w35712
053
Reposted by Rafael M Batista
Aran Nayebi @anayebi.bsky.social · 11/09/2026
CMU is hiring an Assistant Professor in NeuroAI! Come join our awesome NeuroAI community 🧠🤖: apply.interfolio.com/189268
apply.interfolio.com
Apply - Interfolio {{$ctrl.$state.data.pageTitle}} - Apply - Interfolio
02811
Reposted by Rafael M Batista
Gillian Hadfield @ghadfield.bsky.social · 11/09/2026
I spoke to TIME about the Hugging Face incident and AI agents. If you said we're building new members of a group, our group, you'd build them differently than you're building them now. Alignment is not just an engineering problem. It's fundamentally institutional. buff.ly/YO18tlS
152
Reposted by Rafael M Batista
New_ Public @newpublic.org · 10/09/2026
Thanks to AI-powered tools, coding is more accessible than ever, but that doesn’t mean builders have decentralized tech know-how. Grey Area is offering an online course called “Beyond Big Tech” for community builders and designers building alternative tech futures! Apply by September 24th
grayarea.org
Beyond Big Tech: Building Alternative Futures for Personal Data and Infrastructure
Explore data sovereignty, self-hosting, and the decentralized web while gaining hands-on experience with tools for building community networks and alternatives to Big Tech.
054
Reposted by Rafael M Batista
Timnit Gebru @timnitgebru.blacksky.app · 09/09/2026
You guys should read More Everything Forever by Adam Becker about these people. www.hachettebookgroup.com/titles/adam-...
hachettebookgroup.com
More Everything Forever
This "smart and wonderfully readable" (New York Times) exposé shows why Silicon Valley’s heartless, baseless, and foolish obsessions—with escaping death...
332443
Rafael M Batista @rafmbatista.bsky.social · 10/09/2026
#Debunkbot by @tomcostello.bsky.social for #AISafety debunkbot.com?topic=ai-saf...
debunkbot.com
AI Safety
Ask any question. Get a strong explanation.
011
Reposted by Rafael M Batista
Melanie Mitchell @melaniemitchell.bsky.social · 09/09/2026
What, an Anthropic researcher states (without evidence) that there is a 10% chance that AI will cause human extinction? Incredible, we have never before heard such a claim!
2027390
Reposted by Rafael M Batista
Semble @semble.so · 09/09/2026
Found a cool link but don't know which collections to add it to? We now recommend collections for any given link 🪄 Whether you've got a ton of collections to search through or you want to surface related open collections, organizing and collaborating just got a whole lot easier!
1192
Reposted by Rafael M Batista
NBER @nber.org · 04/09/2026
New NBER Book released: The Economics of Transformative AI www.nber.org/books-and-ch...
0134
Reposted by Rafael M Batista
Econometrica @ecmaeditors.bsky.social · 31/08/2026
Can ML generate theoretical insight, not just predictions? Our procedures turn predictive models into anomaly generators, producing minimal cases existing theory can't explain. We apply it to revisit choice under risk. @sendhil.bsky.social @asheshrambachan.bsky.social buff.ly/F0ouzAV
023