235 | OpenClaw: What Happens When AI Goes Unsupervised?

What happens when you let AI agents run completely unsupervised — no rules, no guardrails, no humans in the loop? In January 2026, we all found out. And it got weird.Josh and Will are back to breakdown the Moltbot (now OpenClaw) saga: the viral social experiment that dropped 1.4 million AI agents into a Reddit-style platform and just... let them go. We're talking agents spending $150 of someone's tokens, bricked phones, a Wired journalist secretly infiltrating the platform posing as an AI, and a whole lot of unanswered questions about what this all actually means for the future of AI autonomy.But this episode isn't just about Moltbot — it's about the bigger lesson underneath it all: the human still needs to be in the loop. Josh shares a wild story from a school presentation that stopped a room cold, and they dig into why that principle matters now more than ever as AI agents get more capable and less predictable.Head over to our website at ⁠⁠⁠hitechpod.us⁠⁠⁠ for all of our episode pages, social links, and ways to support us.Need a journal that's secure and reflective? Check out our episodes on the Reflection App, and then sign-up for the App⁠ today! We promise that the free version is enough, but if you want the extra features, paying up is even better with our affiliate discount.Ever wanted to create detailed walkthroughs in the easiest way possible? Check out our episode on Scribe and all that it can do for your training needs, SOPs, or troubleshooting docs.Build a world limited only by your imagination in Topia! A virtual world-building tool built to bring you and any of your virtual guests together. Interested in signing up and learning more? Reach out to us or Topia and let them know we sent you!