An interview with Sam Altman and Greg Brockman on OpenAI's restructuring, cutting Sora, “personal AGI”, Anthropic's “fear-based marketing” for Mythos, and more
The last two standing - on everything — Sam Altman and Greg Brockman came on Core Memory together for a ten-year look back at OpenAI.
Context & Ripple Effects
OpenAI’s prior executive interviews emphasized a broad consumer and infrastructure agenda, while the recent Brockman coverage focused on research direction and the choice between text-centric systems and world models. This appearance turns those themes into an explicit portfolio and organizational discussion.
The company is also positioning “personal AGI” against Anthropic’s messaging around Mythos, making product strategy and safety framing part of the competitive contest for user and institutional trust.
First-order effects
- OpenAI will stop developing Sora and manage a workforce restructuring, concentrating leadership attention and internal resources on the company’s remaining priorities.
- Altman and Brockman publicly sharpen OpenAI’s product narrative around “personal AGI” while directly contesting Anthropic’s framing of Mythos.
Second-order effects
- The Sora decision gives rival video and world-model efforts an opening to compete for developers, creators, and talent that had viewed OpenAI as a likely long-term platform provider.
- Anthropic and other frontier labs face pressure to distinguish whether their public safety messaging is product differentiation, risk communication, or both; OpenAI is explicitly contesting that framing.
Third-order effects
- If leading labs continue pruning high-profile modalities while concentrating on assistant-like products, the frontier-model market may become more focused on owning the primary user relationship than on offering every generative-media surface.
- Competition over increasingly personal AI systems is likely to make governance, trust, and the language used to characterize model risks more central to product positioning—not merely policy debate.
The trend: Frontier AI labs are increasingly pairing organizational and product consolidation with a fight to define which model of personal AI users should trust.