OpenAI CEO Sam Altman Weighs In on AGI, DeepSeek, and Working With the Trump Administration

Speaking on The Times Tech Podcast in London, OpenAI Chief Executive Sam Altman reflected on the rapid acceleration of artificial intelligence since the launch of ChatGPT, predicting that individual productivity will expand dramatically over the next decade. Discussing OpenAI's recent launch of its Deep Research feature, Altman estimated that the tool can already handle roughly 5 percent of economic tasks, particularly in knowledge-based professions. While acknowledging that the labor market transition could bring short-term pain over the next five to ten years, Altman expressed long-term optimism about AI's impact while noting the difficulty of defining artificial general intelligence (AGI).

Altman also addressed the recent emergence of Chinese rival DeepSeek, praising its product choices while explaining that the cost of a given level of AI intelligence drops tenfold every year—a trajectory far steeper than Moore's Law. Despite efficiency improvements, Altman emphasized that massive computing infrastructure remains necessary. He pointed to OpenAI's planned $500 billion Stargate project in the United States and expressed enthusiasm for potentially building similar large-scale AI infrastructure in Europe.

On politics and regulation, Altman advocated for international safety standards modeled loosely after the International Atomic Energy Agency (IAEA) rather than framing global AI development purely as an arms race. He praised President Donald Trump's approach to permits, power generation, and domestic manufacturing, contrasting it with the Biden administration's stricter infrastructure stance. Additionally, Altman briefly commented on his strained relationship with Elon Musk, stating that while he still respects Musk's historical accomplishments, the tech mogul "has really changed" amid ongoing legal disputes.