No spam, ever. We'll never share your email address and you can opt out at any time. Already a subscriber? Log in
Reinforcement learning and autonomy researcher
"Our most capable agent autonomously resolved 9 of 353 open Erdős problems at the per-problem cost of a few hundred dollars, proved 44/492 OEIS conjectures, and is being deployed in combinatorics, optimization, graph theory, algebraic geometry, and quantum optics research." Ni…
Excited to share our paper! Due Process on Hold: A Queueing Framework for Improving Access in SNAP arxiv.org/abs/2605.15165 Millions of Americans interface with the social safety net via call centers that are too congested. In Holmes v. Knodell, bad operations = procedural due…
pip install py123d 3,300 hours of multimodal driving data at your fingertips (Physical AI AV + nuPlan + … + can add custom datasets)! 📄 Paper: arxiv.org/abs/2605.08084 ⭐️ GitHub: github.com/kesai-labs/py123d Monumental effort led by @danieldauner.bsky.social to accelerate open…
New RL variance reduction estimator just dropped: arxiv.org/abs/2605.19235
Given the concerns about AI writing, be aware that Pangram has a super high false positive rate! It frequently rates my personal writing as 100% LLM written!
Engineers now think they don't need managers now that they have AI. Managers think they don't need engineers now that they have AI. Fun dichotomy.
Cannot believe anthropic managed to trick people into handing it 50 billion dollars
There is a type of AI person who gets real joy from stressing people out about the rate of progress
There is of course a temptation to submit AI slop code but that’s not new, the temptation to rush and do bad work is always there. What’s new is a belief over whether the code can be trusted without reading it
If AI hadn’t come around, what excuse tech companies would be using for the layoffs. Oil prices too high?
There is a claim that the OpenAI model was a "general model" rather than a specialized one. But these generalized models are presumably trained on tons of synthetic math data so it's also hard to think of them as "general"