What if your AI assistant could negotiate your next car purchase, trade prediction markets, manage your inbox, and research crypto… while you sleep?

This week we break down the insane rise of ClaudeBot → Moltbot → OpenClaw, the open-source AI agent that rocketed past 200,000 GitHub stars and sparked a massive wave of autonomous agents almost overnight.

But of course, crypto had to crypto.

Along the way, the project triggered a triple rebrand, a $16M scam token spun up by opportunists, and a security mess that unfolded in real time as attackers hunted for poorly-secured instances and exposed keys.

We also get into the bigger picture: as AI agents start doing real work for people (and eventually paying bills, trading, and moving value), crypto becomes the natural payment rail. Permissionless. Programmable. Always on. Which is exciting… and also a whole new playground for scammers.

We cover what happened, why it matters, and what you should do if you’re experimenting with agents: lock it down, separate machines/accounts, protect your keys, and don’t chase random tokens.

Show notes and links: http://badco.in/804

Leave a comment with the best OpenClaw tutorial you’ve found — we’ll dig in.