scripod.com

AI Regulatory Capture Bet: Why OpenAI & Anthropic Stopped Competing on Tech

More or Less

21 HOURS AGO
More or Less

More or Less

21 HOURS AGO
The podcast dives into the shifting landscape of AI, where the hosts debate whether the model quality wars are over and if the real competition has moved to politics and narrative. They also explore the implications of new open-weights models, the hype around AI hardware, and the cultural trends shaping Silicon Valley.
The hosts argue that the AI model arms race is effectively over, with open-source and cheaper alternatives like GLM 5.2 matching frontier labs in quality, making premium models less cost-effective. They suggest that companies like OpenAI and Anthropic are now focusing on regulatory capture and narrative control rather than technological breakthroughs, as the real competition has shifted to Washington. The discussion also covers OpenAI's new smart speaker, which is dismissed as an unoriginal Alexa-like device, and the broader trend of Silicon Valley becoming like Hollywood, where valuations depend on narrative and star power. The episode concludes with lighter topics, including the improvements in real-time voice AI, the unreliability of biological age tests, and the social dynamics of Hinge's new 'Friend's Take' feature.
07:24
07:24
Open-source models have caught up to frontier labs.
19:22
19:22
AI competition has become political rather than technological
24:15
24:15
Policy proposals are political gestures, not plans.
31:49
31:49
Valuations depend more on narrative than revenue.
48:33
48:33
Biological age tests are clout scores for the body.