Sam Altman: “The world should trust” us on AI safety

After a week of dire warnings, the CEOs of OpenAI and Anthropic toned down doom messages at the Dreamforce conference.