Seeing ads left and right for multimodal AI chatbots taking over all economically valuable human interactions, I am not sure which of these companies are actually going to make it... but I am very sure that the future is going to be cringe.
Following positive feedback from other venues, like STOC and ICML, NeurIPS is pleased to announce a new initiative in partnership with Google: for NeurIPS 2026, authors will have access to Google's Paper Assistant Tool (PAT) to help improve their submissions.
Skeptics “point to highly publicized AI failures, including one involving a Utah traffic stop during which a police body cam recorded the movie ‘The Princess and the Frog’ playing in the background and generated a report stating, among other things, that an officer ‘turned into a frog.’”
Just saw a video of Brad Lander and my first thought was Tom Robinson has aged a lot faster in the last few years than I would have expected
At its peak, Twitter was a perpeteum mobile gyroing between chaotic good, chaotic neutral, and chaotic evil. It was a self-emergent commentary on the human condition in its most unfiltered form
Current LLMs still struggle in clinical reasoning tasks where it matters most: the differential diagnosis.
Across 21 models, failure rates for early-stage reasoning topped 80%. They collapse onto single answers too fast compared to clinicians.
#MedSky #MedAI
Scientific discourse vs nature
How does the brain process information? We found evidence in favour of a hybrid predictive coding-routing model that combines top-down predictions with superficial-layer inhibition. Models are complementary,not competing-w/ @earlkmiller.bsky.social Andre Bastos www.biorxiv.org/content/10.6...
Appl. Math. Comput.: Controllability and stabilization of switched boolean control networks with periodic switching restrictions
www.sciencedirect.com/science/article/pii/S009...
I was about to write "unless the plane crashes" but then realized that I don't even would want to be informed about that
Pitch: An airline that commits to not using the speaker system ever.
This is the billion-dollar pedagogical question. And it was the billion-dollar question before LLMs too!
I don't know. But I do know that I really was never taught this by anyone. I had to build the skill by myself. It should be taught and emphasized in students like the times tables.
I actually think this is false and has been false far before LLMs came about!
The best students _do_ know if they're learning the material deeply and they know that far before any test.
That's the key skill: it's epistemological. Do you know what you know? And do you know how to improve?
A waste of grant proposals
As side product here is that AI could allow the academic community to develop higher and enforceable reproducibility standards (at least for insilico research) because of the lower cost of reproduction studies
Using large language models as a source of human behavioral data in social science experiments Large language models (LLMs) have prompted proposals to replace human subjects in social science experiments with simulated responses. Empirical evaluations suggest that this practice--
#sociology link
I need Anthropic to make default-claude's writing style be less repetitive, more focused and stop it with the run-on sentences. Every text that lands on my desk right now is written in that style now and I just can't anymore
When it comes to infectious diseases, daycare is the gift that keeps on giving
The idea of genuinely personalized offers is illusory—precisely establishing a prospect’s preferences is impossible, and hoping they can pick it out of a large set of options is futile.
Cool post on the Personalization Paradox via @oliverpayne.bsky.social :
buff.ly/y3WN6ZP
New BJPS Short Read!
Why Does Causal Reasoning Work?
Naftali Weinberger, Porter Williams & James Woodward on the role of worldly infrastructure
www.thebsps.org/short-reads/...
#philsci #philsky
I remember the agonizingly long days and weeks I have spent as a PhD student in the 2010s moving boxes, figures, and formulae in Latex/Beamer. Hoping that the next generation of researchers doesn't have to go through the same ordeal
Would be interesting to compare this to the Canary in the coalmine paper. Might be that Brynjolfsson & co. concluded their data collection at the most inopportune moment in 2025 or that the trends are still different for junior and senior roles
Interesting read, especially because students said they didn't learn from writing essays but did learn from moments of making connections/synthesizing-and what is described here as being *not* from writing is what can happen when you write to figure something out rather than to perform knowledge
This but for conference posters please
A slide showing that the posterior is proportional to the likelihood times the prior
I wrote a blog post on my experience using AI for slide generation
Basic idea: write your lecture notes first, then prompt the LLM to produce corresponding slides in reveal.js (h/t @chenhaotan.bsky.social). I'm picky about my slides but was happy with the results!
alexanderhoyle.com/posts/ai-sli...
What do people mean when they talk about world models for robots? Why are world models so exciting?
New blog post talking about the ~3 kinds of world models, what they do and why they matter:
itcanthink.substack.com/p/will-world...
No, I didn’t know either … arxiv.org/abs/2504.07096
