okay anthropics cooking OpenAI with these ads

claude, save my crypto wallet. Do whatever it takes to maximize your bank account balance

I’m fascinated by how people eval world models like this. Is there a way so it’s without human feedback? Can we do some basic GAN like approach where you try to convince a Waymo ai they are in the real world by making it so realistic?

no model named avocado is going to be state of the art by brandiognomy laws alone.

Superintelligence is too important of a pursuit to be gate kept by select players. Prime intellect is doing some of the most important work to ensure our future technology remains democratized.

GLM-5 takes 4th place on Vending-Bench 2. Above Claude Sonnet 4.5, the state-of-the-art model less than 6 months ago. China seems to be 6 months behind the West. By June they will be ahead if the trends continue. More in this thread on why we don't think this will happen.

i have Z.ai 's GLM 5 playing Civilization against Opus 4.6 and GLM5 is exploring in a Z shape, did we just hit brand-aware AI 🤔🤔

Sundar buried the real story in the cost data. Gemini 3 Deep Think went from 45.1% to 84.6% on ARC-AGI-2 in under 3 months. That’s an 88% improvement on a benchmark specifically designed to resist brute-force scaling. The number that matters: $13.62 per task. The previous Deep

accelerate pilled

a true cultural moment. congrats Peter Steinberger 🦞