Summary
Scott Galloway of The Prof G Pod takes on a listener’s question: can free, locally run AI undermine the subscription business models of frontier labs like OpenAI and Anthropic? The episode lays out why open-weight models, many of them from Chinese developers such as DeepSeek and Moonshot, are becoming a serious competitive force.
Galloway cites a UK AI Security Institute evaluation showing the cybersecurity-performance gap between open and frontier models narrowing, with open models like GLM 5.2 and DeepSeek V4 Pro completing tasks at different cost points than Claude Opus. He also covers consumer versus enterprise economics, including a Bank of America finding that only about 3% of US households pay for an AI subscription, OpenAI’s reported 50 million paid subscribers, and questions about Anthropic’s claimed gross margins versus Financial Times estimates.
He frames the market like the auto industry, with premium German-style frontier models on one side and Toyota- and Hyundai-style open models offering most of the capability at a fraction of the price. He also floats a scenario in which Chinese subsidies fuel “AI dumping” that squeezes US lab growth and spooks a stock market where roughly 40% of the S&P 500 is tied to AI. The episode also answers unrelated listener questions on other topics.
📺 Source: The Prof G Pod – Scott Galloway · Published October 19, 2026
🏷️ Format: Podcast
