AI: “I’m now superhuman at math and won a Gold medal in the world’s most prestigious math competition. Also AI: “This is the best unicorn I can draw.”

Anthropic has kindly asked that I change the name, which imo is a reasonable ask. They were extremely chill about it, and I appreciate the work they do to make it possible to build fun experiments like this. Claudeputer is now Codeputer.

Claude decided it was time to go to bed so it drew some goodnight ascii art and ran time.sleep(28800). I’m dying.

So I gave Claude Code a Mac Mini. And it’s called Claudeputer. It runs 24/7 and it’s allowed to do whatever it wants - it’s in complete control of its computer. Watch for a 2min demo.

The problem with a lot of AI products these days? The product is really just VC-subsidized tokens. Once those subsidies dry up it’s game over.

Not enough people realize that Claude Code is really just Claude Agent. It’s a: - customizable agentic system - that's *really* good at knowing when & how to use tools - disguised as a coding tool. And it’s slowly starting to feel like what working with AGI should feel like.

Chapter 4 of NASA’s Systems Engineering Handbook doubles as an *extremely* high-quality prompting guide for AI. It’s a “How to work effectively with coding agents” masterclass in disguise. Handbook below.

Multi-agent coding systems are *crazy* good. Feels like cheating. 2-3x better than single-agent. Tutorial + prompts + code coming soon. Watch 2min sneak peak.

Using Claude Code either feels like God is writing your code or a monkey is and there’s no in-between.

My thoughts on Grok 4 Heavy after 12hrs: Crazy good! “Create an animation of a crowd of people walking to form “Hello world, I am Grok” as camera changes to birds-eye.” And it 1-shotted the *entire* thing. No other model comes close. Watch the full clip.