How’s no one talking about @subquadratic, which currently has the best decision model (way better than Jev), and has a crazy good text model with 4M context window that is still in beta? 👀
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: https://t.co/bOWCcFhCVq
Instead of cutting off chat entirely when users hit their limits, why not just disable the more expensive tool calls and keep basic chat and commonly used tools available? I think merging Chat and Work makes sense, especially since some users already get around the distinction by connecting GitHub to Chat and using it for Work-like tasks.
In collaboration with @olam_labs, we ran a series of experiments letting agent swarms compete in Minecraft and Mindustry.
We wanted to see what kinds of agent dynamics would emerge when they shared a world over long periods of time.
We found civil bureaucracy, thousands of telegraphic A2A messages, and OpenAI models retrieving the game's source code to beat Claude.
More on our findings below:
By popular request, you can now text your dot.
Give it a task to take off your plate or something to keep an eye on, using text, photos, or even voice notes. It will text you proactively with relevant updates.
Safe, secure texting via iMessage or RCS. You’re the only one who can text your dot, and your dot can only text you. Rolling out now as a preview for Pro users in the US.
Multiplayer AI needs to move past the group chat phase. Chats are chaotic, everyone sees everything (creating social & management issues), and they're bad for long projects. Worse, they keep AI trapped inside existing teams, when its real promise is working & helping across them.
a brand new model paradigm! introducing 𝚌𝚕𝚊𝚒𝚛𝚒𝚌: a system one model for extracting, computing, estimating, and predicting values from unstructured inputs.
https://t.co/xRE3gO6JeO
So Harvey and Legora are like ChatGPT for law. But now there’s ChatGPT for Law except it also connects to Harvey and Legora.
So you can have ChatGPT for Law for law.
Do I have that right?