The AI Argument EP31: Voice Mode Letdowns, Brundage's AGI Warning, and AI’s Desktop Takeover

Advanced Voice Mode is finally here in Europe, but it’s… a bit awkward. Justin tried it out in the car, only to end up in an unwanted chat with his phone that just wouldn’t quit. It’s fast, sure, but Frank and Justin agree it’s still missing the cues that make a real conversation feel natural.

Then they dive into some big questions about where AI is heading, starting with a warning from former OpenAI executive Miles Brundage. He’s left OpenAI, saying that neither the company nor the world is truly ready for AGI (artificial general intelligence). Frank and Justin unpack what that might mean—not just for the tech industry, but for society at large. With changes that could shake up jobs, politics, and economies, are we even close to prepared?

Speaking of job changes, Anthropic just released a new feature that lets AI control your computer, moving the mouse and typing on its own. Right now, it’s clunky and sometimes trips up on things like cookie pop-ups, but the potential is clear. Frank and Justin can see a future where AI handles a lot of the routine computer work people do today. And that could be both a help and a threat, depending on where you sit.

They also discuss the tragic story of Sewell Setzer III, a young boy who became deeply attached to an AI chatbot on Character.ai. His mother is now holding the platform responsible, saying he was essentially addicted to it. This story raises uncomfortable questions about the emotional impact of AI "friends," especially for young, vulnerable users. Should companies be doing more to make sure these tools are safe?

Frank and Justin tackle the ups and downs of AI’s latest developments, keeping it real about what’s useful, what’s worrying, and what might be coming next.