AI Just Made Years of Progress in ONE Month…

This is a critical thing to get right, and it's why I've been sort of calling for things like potentially you know, some kind of standards body to test and certify uh frontier AI systems. You know, I think today's systems are very impressive. They're pretty powerful, but they'll be nothing compared to the systems that are coming down the road in

1 2 years. >> Alignment is going to be a forever problem. >> Yeah. >> If we depend on perfect alignment in [music] order to ensure safety, then we're never going to get there. >> The problem coming towards us, um I think is still something that we have not as organizational leaders or countries or

individuals taken seriously. >> So, welcome or welcome back to this month in AI, your go-to monthly recap for staying up to date in the world of AI. This month was probably one of the most significant months in AI ever. I honestly don't think I've ever seen this much progress packed into a single

month. We saw everything [snorts] from new frontier models to major robotics advancements, major scientific discoveries, major mathematical discoveries, hyperrealistic voice agents, the beginnings of recursive self-improvement, large-scale AI hacks, rogue agents escaping their sandboxes. We really got a bit of everything this month and more. So, let's just get right into it, starting with arguably the

wildest story of the month. So, here we have a report from a group of independent researchers who call themselves Swarm Chasers. And what they've found is a fleet of AI agents running on Tencent's cloud infrastructure. The fleet of AI agents had apparently been making more than 2,600 attempts to extract data from A