FULL INTERVIEW: Sam Altman Responds to Anthropic’s Attack Ads, Live on TBPN

Summary

Sam Altman joins TBPN to discuss OpenAI's launch of GPT-5.3-Codex, the shift toward managing AI agents, and the importance of compute, chips, and power. He addresses AI's impact on software and SaaS, social/video products, OpenAI's advertising stance, and competition with Anthropic. The conversation also covers bottlenecks in AI infrastructure, fabs, and the broader economy.

• OpenAI launched GPT-5.3-Codex, emphasizing coding, computer use, and mid-turn steering.
• Altman expects users to manage teams of AI agents, making orchestration tools key.
• Compute power is described as the new oil, with chips the current bottleneck and fabs/wafer capacity needing more funding.
• Software and SaaS are not dead but face volatility, sell-offs, and transformation.
• OpenAI is testing ads, plans to avoid ads in the LLM stream, and disputes Anthropic's attack ads.
• Altman discusses social agents, Sora/video generation, Disney IP, and voice/hardware.
• He comments on AI research labs, model IQ limits, space data centers, and OpenAI's public scrutiny.