This is major news in the AI world 💣
Andrej Karpathy has joined Anthropic 🚀

He was one of the 11 founding members of OpenAI.

Today, Andrej Karpathy announced that he is joining Anthropic’s pre-training team. Even more interestingly, he will be building a new team focused on accelerating pre-training research from within — using Claude itself.

As he wrote: “The next few years at the frontier of LLMs are going to be especially formative.”

➡️ Read that sentence again.

He led the computer vision team for Tesla Autopilot.
He worked at OpenAI twice.
He founded Eureka Labs — a project aimed at rethinking education with AI.
He helped popularize the term “vibe coding.” ‼️

➡️ He could have gone anywhere. He chose Anthropic.

So the real question is: how do you track the AI race?

Funding rounds?
Model names?
Benchmark tables?

People like Karpathy do not always show up in those tables. But every move they make sends a clear signal.

And the signal here is this: pre-training is not over. Large language models have not reached their final form. The most important technical work is happening right now.

Saying “we use AI” is not the same as being able to read these signals.

If you are choosing vendors, making architecture decisions, or building AI into your business today, the technical landscape is changing every day.

You cannot understand it only through benchmark tables.

You need to read the signals.