OpenAI's TBPN Acquisition, AI Agent Burnout, and Why Texting Is the Future of AI

Alex Heath and Ellis Hamburger kick things off by dissecting OpenAI's acquisition of TBPN—a move that raises thorny questions about media ownership and corporate consolidation in the AI space. They dig into the tension between maintaining editorial independence and being owned by a major player in the industry you're supposed to cover critically. It's the kind of conflict that haunts tech journalism, and they don't shy away from the awkwardness of it.

But things get darker when they pivot to a genuinely unsettling trend: the epidemic of burnout among AI researchers and founders. People in this space, they note, seem to be grinding themselves down to nothing—working inhuman hours, sacrificing sleep and relationships, all in service of this gold-rush mentality. It's not just individual overambition; there's something structural about the AI industry that seems to demand human sacrifice in a way that feels almost cult-like.

Then Marvin von Hagen, CEO of Poke, joins the conversation and brings a refreshingly grounded perspective. Poke is essentially an AI agent that lives in your text messages and can actually *do things*—pull info from your email, check your calendar, execute tasks—without requiring you to jump between apps or use some clunky interface. It's the kind of seamless integration that feels obvious in retrospect but is apparently difficult to execute well. Marvin explains why he believed in AI agents early when everyone else was still obsessed with chatbots, and he makes a compelling case for why texting is actually the perfect interface for AI. It's how humans naturally communicate, and it eliminates the friction of learning yet another app.

What's particularly interesting is how Poke approaches the security and automation question—the stuff that terrifies people about AI agents. Rather than going full "autonomous AI that might send an email without you knowing," they've built in safeguards and clarity around what's happening. Marvin also touches on his lawsuit against Meta, a legal battle rooted in principles about AI ethics and corporate accountability that feels almost quaint given how fast the industry is moving. The conversation highlights the gap between Silicon Valley's move-fast-and-break-things ethos and the reality that, sometimes, things shouldn't be broken.