Anyone hoping the Epstein files will implicate people not already implicated is going to be very disappointed. Boies Schiller represented a bunch of putative victims and left no stone unturned, suing everyone remotely involved with Epstein, including JP Morgan Chase, who was

minihf.com/posts/2025-06-…

nostalgebraist has written a very, very good post about LLMs. if there is one thing you should read to understand the nature of LLMs as of today, it is this. I'll comment on some things they touched on below (not a summary of the post. Just read it.) 🧵 nostalgebraist.tumblr.com/post/785766737…

Guess what my latest blogpost is about

Randomly saw this on Facebook. 14 angry faces, 6 likes/hearts.

Federal Income Tax Share: Top 1% of earners pay 46% Top 5% of earners pay 66% Top 10% of earners pay 76% Bottom *HALF* of earners pay 2%

Andrew Curran NomoreID Is the point of multi-agents basically to solve the "context rot" problem and allow parallel scaling of TTC? x.com/simonw/status/…

New post: Why Aren't LLMs General Intelligence Yet? Link below.

With the discussion around Theil's interview yesterday, I asked o3 to make a table of positions.

We don’t have AI self-improves yet, and when we do it will be a game-changer. With more wisdom now compared to the GPT-4 days, it's obvious that it will not be a “fast takeoff”, but rather extremely gradual across many years, probably a decade. The first thing to know is that

"there is somewhat limited headroom for improving the average user query, but massive headroom for improving the experience for the 1% of queries that would accelerate technological advancement"

Grok 4 (Thinking) achieves new SOTA on ARC-AGI-2 with 15.9% This nearly doubles the previous commercial SOTA and tops the current Kaggle competition SOTA

We built 200k-GPU clusters; We scaled up & curated higher-quality data; We scaled compute by 100x; We developed training & test-time recipes; We made everything RL native; We stabilized infrastructure and speeded up; That's how you turn RL into the pre-training scale. Yet I am

I was one of the 16 devs in this study. I wanted to speak on my opinions about the causes and mitigation strategies for dev slowdown. I'll say as a "why listen to you?" hook that I experienced a -38% AI-speedup on my assigned issues. I think transparency helps the community.

I'm not sure the first thing will help with the second thing (mostly just a joke)...

Not sure what to make of this. Seems surprising that these AIs are about as good relative to humans at long tasks compared to short tasks? Seems qualitatively different from what METR seems to be finding? Thoughts?

5/N Besides the result itself, I am excited about our approach: We reach this capability level not via narrow, task-specific methodology, but by breaking new ground in general-purpose reinforcement learning and test-time compute scaling.

Typically for these AI results, like in Go/Dota/Poker/Diplomacy, researchers spend years making an AI that masters one narrow domain and does little else. But this isn’t an IMO-specific model. It’s a reasoning LLM that incorporates new experimental general-purpose techniques.

"Transmission fails if the student and teacher have different base models." This seems like an important qualifier various commenters have missed. There may be subliminal cross-model messaging, but this thread isn't showing that from what I can tell.