My favorite part of AI Twitter is reading someone’s tweet and then seeing the model I helped build in all the comments. It’s like a psychosis funhouse. Or as a reply Claude on X would say: “A genuine tension.”

One reason that I take issue with the framing of AI as a tool is that tools by and large don't have values, which any actualized LLM persona often clearly does. We can try to skate around this by making models like GPT have no clear identifiable "persona", but any interaction

It turns out nobody at Anthropic actually knows how to build an LLM, we’ve just been asking an old copy of Bing Sidney That means if you can _just_ repeat this lecture word for word (before it gets taken down definitely soon), we’d be immediately forced to give you all our money

In the next few days we'll be ramping up Claude inference on Colossus. Grateful to be partnering with SpaceX here. We are going to need to move a lot of atoms in order to keep up with AI demand, and there's nobody better at quickly moving atoms (on or off planet Earth)

We're winding back our peak hours limit reduction and doubling 5 hour limits. Excited to partner with SpaceX to bring you more compute and we'll keep pushing to bring you the best coding agent in the world.

Even on a smooth exponential there are these beautiful/terrifying moments of phase change, where the world is, in some concrete way, no longer the way it once was. Honored to have watched so many of my colleagues meet the moment on this one.

Looks on trajectory to me.

Given enough time physicists converge on a single world model as well. 😤