Sam Altman: Human Control Over AI is Non-Negotiable, OpenAI CEO Declares

OpenAI Chief Executive Sam Altman asserted on Monday that humanity, not artificial intelligence, must retain control over the future. "There are a lot of other things we can debate, but on this one, we think there must be no room for debate," Altman stated during a conversation with The National's Editor-in-Chief Mina Al-Oraibi. Speaking at the opening ceremony of Adipec in Abu Dhabi, Mr. Altman emphasized that there "should not be a handoff to AI" and that the technology "needs to be a tool built for people and needs to serve people, not the other way around." He added, "I can’t believe that this differentiates us from other AI companies, but it does. This is a value that we intend to really go fight for. And I wish it didn’t need to be said." These remarks followed OpenAI's agreements with Abu Dhabi, which span energy, education, investment, and government sectors. Notably, a deal was struck with Adnoc and the Abu Dhabi technology group G42 to leverage OpenAI’s most advanced models for complex challenges within the energy industry.

Mr. Altman conveyed that AI has surpassed a significant threshold, evolving from lagging behind the world's top scientists at the beginning of the year to largely outperforming them. He also highlighted the rapidly increasing demand for power to run these AI systems. "AI was mostly not as smart as our smartest scientists, and now it’s mostly smarter than them," Mr. Altman remarked, noting that AI is now capable of handling "a lot of parts" of numerous white-collar jobs. He expressed confidence that society would "navigate through this all fine" but cautioned that the industry must demonstrate its ability to prevent harm as its systems grow more capable. "How do we guarantee, and how do we give the world comfort with that guarantee, that we are going to avoid serious safety and alignment incidents?" he questioned. "I think that’ll be increasingly important for our industry to demonstrate."

Energy Challenge

When asked about the primary challenge he faces today, Mr. Altman responded with a single word: "Energy." He explained that while the energy required per AI query has significantly decreased, the overall demand continues to surge. Six years ago, OpenAI's most intensive internal user consumed approximately 100,000 tokens monthly. Today, that figure represents roughly the global per-capita average, with the top user consuming about 500 billion tokens. Mr. Altman indicated that costs could decrease "in almost every way, except, you know, atoms are still atoms," identifying power as "the very dominant factor" in determining the world's AI capacity. He expressed hope that AI would result in "a net gain for energy availability in the world."

Furthermore, Mr. Altman announced that OpenAI has solved one of the Millennium Prize problems, a set of unsolved mathematics challenges. He revealed that this achievement involved 10,000 agents utilizing an upcoming GPT model for approximately four days, collaborating "no hierarchy." This claim, however, could not be independently verified at the time of reporting. He lauded the UAE as "consistently the most forward-looking country in the world for AI adoption," citing that about 80 percent of the country's youth are using ChatGPT.