Chinese AI Model GLM 5.2 Beating Frontier Models | Meta Glasses, Polymarket Scandal, and More AI Talent War
About this episode
The episode discusses Meta's new smart glasses, prediction markets, the performance of AI model GLM 5.2, and the ongoing AI talent war.
The quad is back. Dave and Brit rejoin Sam and Jess fresh off a More or Less fan encounter at the Zurich airport. It was quite the week for Meta as the squad breaks down the company’s new line of glasses featuring Kylie Jenner (or, in Jess’ words, Kendall), and Brit makes the call that this is finally the watershed moment for women buying smart glasses ( The men seem to be the early adopters, Dave wore his in the Sistine Chapel, and Sam owns 10+ pairs.). Then they get into Meta’s prediction market: a genuinely interesting information machine or a very efficient way to addict twenty-somethings to fake gambling? In true MOL fashion, the squad is split. Sam is all in, Dave is worried it’ll work too well, and Brit thinks it’ll all click when you tie it to The Bachelor. Dave also ran GLM 5.2 on eight Nvidia B200s with some hacker friends and came away shaken. Open source is now matching, if not outperforming, frontier models at 10x the speed, raising uncomfortable questions about the enormous bets sitting on frontier labs today. Remember last year’s AI talent war? It’s back, and Jess argues it’s really an early indicator of investor insecurity. Finally, in pop culture corner: Cannes…
More episodes of More or Less
- AI Regulatory Capture Bet: Why OpenAI & Anthropic Stopped Competing on Tech
- OpenAI Offered the Government a Stake. Is This a Distress Signal?
- Why Identity Verification Matters For The Internet | Vox Americanus, GLM 5.2, Claude Fable 5 Is Back
- Elon Is the Best Financier of Our Generation
- Teamshares IPO: $60M EBITDA, 92 Companies, Zero Exits
Explore listener stats, chart rankings, contacts and more on the More or Less podcast page.