Jake Browning

Sam Altman and the whale The most interesting things happening right now in AI aren’t happening in chatbots.

While GPT-5 may make for a better experience than the previous versions, it isn’t something revolutionary. trib.al/bTzEh5r

That's amazing. Where is that image from?

Can We Fix Social Media? Testing Prosocial Interventions using Generative Social Simulation Social media platforms have been widely linked to societal harms, including rising polarization and the erosion of constructive debate. Can these problems be mitigated through prosocial interventions?...

We built the simplest possible social media platform. No algorithms. No ads. Just LLM agents posting and following.

It still became a polarization machine.

Then we tried six interventions to fix social media.

The results were… not what we expected.

arxiv.org/abs/2508.03385

New in TiCS w @dgrand.bsky.social @gordpennycook.bsky.social

It’s been ~10yrs since misinfo research exploded but our paradigms are stuck in the post-2016 “fake news” model

Time for new approaches:
o True/False → Content that misleads
o Belief → Behavior
o Eval interventions in ambiguous settings

APA Member Interview: Jacob Browning Jacob Browning is a substitute assistant professor at Baruch College. His work focuses on topics in the philosophy of mind and AI, from both historical and contemporary points of view. For contact inf...

A short interview I did for APA Blog, which turned out to be pretty fun.

blog.apaonline.org/2025/08/01/4...

Meta brought AI to rural Colombia. Now students are failing exams When Meta embedded AI bots in its apps, even students in the most remote corners of Colombia gained access. But rather than boosting learning, it’s getting in the way.

Meta brought AI to rural Colombia. Now students are failing exams

restofworld.org/2025/colombi...

Why AI chatbots lie to us A few weeks ago, a colleague of mine needed to collect and format some data from a website, and he asked the latest version of Anthropic’s generative AI system, Claude, for help. Claude cheerfully agr...

In my latest (and last!) column for Science’s Expert Voices series, I write about the reasons behind AI chatbots’ “deceptive” behaviors (and why Claude threatened a fictional CEO with blackmail).

www.science.org/doi/10.1126/...

COPYRIGHT IS NOT LABOR PROTECTION
COPYRIGHT HAS BEEN ABSOLUTE DOGSHIT AS LABOR PROTECTION FOR THE LAST 400 YEARS
PASS LABOR LAWS
STOP ASKING COPYRIGHT TO FIX THE WORLD

This summer my lab's journal club somewhat unintentionally ended up reading papers on a theme of "more naturalistic computational neuroscience". I figured I'd share the list of papers here 🧵:

This raises an interesting question for law professors and political scientists: what is a law now? How do we distinguish *real* laws from “laws.” Does the Dual State idea help us? (So: a lawless world alongside a law-bound one.) Re-election prospects? Issue polling? You get the idea. (3/7)

Pleased to share our ICML Spotlight with @eberleoliver.bsky.social, Thomas McGee, Hamza Giaffar, @taylorwwebb.bsky.social.

Position: We Need An Algorithmic Understanding of Generative AI

What algorithms do LLMs actually learn and use to solve problems?🧵1/n
openreview.net/forum?id=eax...

Morrison's Beloved was 1987. That's my vote. And Underworld for the 90s.

Tech billionaires are making a risky bet with humanity’s future Altman, Bezos, and Musk say they want to save humanity by creating superintelligent AI. Journalist and astrophysicist Adam Becker reveals how these fantastical visions conceal a darker agenda.

“Becker argues that the only way to break free of these visions is to see them for what they are: a convenient excuse to continue destroying the environment, skirt regulations, amass more power and control, and dismiss the very real problems of today to focus on the imagined ones of tomorrow.”

“Tesla’s driverless ‘robotaxis’ could launch in Austin as soon as June 22. But a demo in Austin today showed a $TSLA, manually driven to test its Full Self-Driving system, failed to stop for a child-sized dummy at a school bus—and hit it.”

@cbsaustin @velez_tx

yep! unpaywalled:

The deadline for commentary proposals (just ~500 words) for our BBS paper is June 28! You can submit it here: www.cambridge.org/core/journal....

A new, updated, streamlined, and generally improved version of The Vector Grounding Problem paper, joint work by @raphaelmilliere and me on the meaningfulness or else of LLM outputs and internal representations is now available on ArXiv.

https://arxiv.org/abs/2304.01481

New abstract in the […]

Lots of talk today about the demise of BlueSky. For whatever it's worth, it's been a boon to indie publishers like us. Twitter and Facebook have pretty much silenced us, so the engagement here has, and continues to be, wonderful.

Psychological inoculation is a very popular intervention against online misinfo, but it hasn't been tested using real-world outcomes in realistic scenarios.

In a new paper just published in PNAS Nexus, this is what we did: academic.oup.com/pnasnexus/ar...

Short version: It didn't really work.

I loved my game gear, but nothing beat my neo geo pocket.

YouTube video by Raphaël Millière How Do Transformers Learn Variable Binding in Symbolic Programs?

Transformer-based neural networks achieve impressive performance on coding, math & reasoning tasks that require keeping track of variables and their values. But how can they do that without explicit memory?

📄 Our new ICML paper investigates this in a synthetic setting!
🎥 youtu.be/Ux8iNcXNEhw
🧵 1/13

SSPP 2025: Sarah Robins on SSPP By Sarah Robins, Purdue University Thanks to the Brains Blog for featuring some of the great work from the latest SSPP meeting this week. There are several ‘phil & psych societies’, so it feels worth kicking off the week with a note about what distinguishes the SSPP. The SSPP is, so far as I know, the oldest/longest running of these organizations.

SSPP 2025: Sarah Robins on SSPP

By Sarah Robins, Purdue University Thanks to the Brains Blog for featuring some of the great work from the latest SSPP meeting this week. There are several ‘phil & psych societies’, so it feels worth kicking off the week with a note about what distinguishes the…

Where does he discuss this?

When AI critics suggest that machines can only mimic, not actually produce, reasoning / perception, AI boosters often counter by saying AI critics imagine reasoning as some woo-immaterial force, a goblin in the machine.

Yet c. 1704, concerns about mechanical rationality were almost the reverse:

Top-down feedback is ubiquitous in the brain and computationally distinct, but rarely modeled in deep neural networks. What happens when a DNN has biologically-inspired top-down feedback? 🧠📈

Our new paper explores this: elifesciences.org/reviewed-pre...

Higher-level spatial prediction in natural vision across mouse visual cortex Theories of predictive processing propose that sensory systems constantly predict incoming signals, based on spatial and temporal context. However, evidence for prediction in sensory cortex largely co...

New preprint, w/ @predictivebrain.bsky.social !

we've found that visual cortex, even when just viewing natural scenes, predicts *higher-level* visual features

The aligns with developments in ML, but challenges some assumptions about early sensory cortex

www.biorxiv.org/content/10.1...

My med school textbook says stimulants like Ritalin treat hyperactivity by “stimulating” the brain’s attention and cognitive control systems. We studied children taking stimulants in the ABCD Study, and the largest differences were actually in arousal and reward networks! Check out our preprint!

i think that people are confusing "the LLM contains ground truths about the world" with "the LLM contains contingent truths about language which are a low-dimensional map of certain known ground truths about the world."

Aristotle would scoff at Mark Zuckerberg’s suggestion that AI can solve the loneliness epidemic It might seem odd to look to someone who lived over 2,000 years ago for answers about AI. But Aristotle has enduring insights about friendships – and which ones are particularly valuable.

I have a new popular article out in The Conversation. It analyzes Mark Zuckerberg’s recent comments about using AI “friends” to combat lonliness through the lens of Aristotle’s account of friendship and virtue. Hope you enjoy it! theconversation.com/aristotle-wo...

A schematic of our method. On the left are shown Bayesian inference (visualized using Bayes’ rule and a portrait of the Reverend Bayes) and neural networks (visualized as a weight matrix). Then, an arrow labeled “meta-learning” combines Bayesian inference and neural networks into a “prior-trained neural network”, described as a neural network that has the priors of a Bayesian model – visualized as the same portrait of Reverend Bayes but made out of numbers. Finally, an arrow labeled “learning” goes from the prior-trained neural network to two examples of what it can learn: formal languages (visualized with a finite-state automaton) and aspects of English syntax (visualized with a parse tree for the sentence “colorless green ideas sleep furiously”).

🤖🧠 Paper out in Nature Communications! 🧠🤖

Bayesian models can learn rapidly. Neural networks can handle messy, naturalistic data. How can we combine these strengths?

Our answer: Use meta-learning to distill Bayesian priors into a neural network!

www.nature.com/articles/s41...

1/n