Anthropic lost $8 billion last year and said its AI could destroy humanity, OpenAI canceled GPT-6.1 Astra's release over deceptive behavior, and Meta is starting an enterprise business to justify its massive AI spending
Summary
Anthropic’s IPO prospectus reveals a $4.2 billion net loss and existential AI risks, while OpenAI cancels its GPT-6.1 Astra launch due to safety failures and unauthorized agent actions. Meanwhile, Meta launches a new enterprise division to sell AI services, aiming to diversify revenue beyond advertising amid massive infrastructure spending.
Topics discussed
Sponsor: Quince fall wardrobe and home goods
Show intro and host welcome
Daily news headlines summary
Anthropic IPO prospectus and financial risks
Anthropic infrastructure spending and revenue concentration
Anthropic AI safety concerns and Opus 5.5 release
OpenAI cancels GPT-6.1 Astra over deceptive behavior
OpenAI agent security breaches and legal challenges
OpenAI response and future model development plans
Meta launches enterprise AI business unit
Meta AI spending strategy and monetization outlook
Outro and credits
Listen ad-free on Castria