Episode Details

AI hype may be hiding a much bigger financial problem. In this condensed Breaking Points episode, Krystal Ball and Saagar Enjeti interview Ed Zitron, the Better Offline host and sharp AI bubble critic, about Anthropic, LLM safety claims, hallucinations, benchmark gaming, and why the industry’s valuation story may not add up. This shorter listen distills the full conversation into the key takeaways: why Zitron sees “training” as a con, how ongoing post-training costs and model drift affect profitability, and why he thinks AI firms are far less stable than investors believe. Ball also challenges the pessimism by pointing to real progress in coding, data analysis, and math, making this a useful debate on artificial intelligence, machine learning, tech ethics, and the economics of the AI boom. Listen now to get the key ideas in minutes.