Altman tells the G20 no child will ever out-think AI — the electricity analogy is the part worth keeping

Sam Altman used a ministerial stage to make a claim about the permanence of the machine's advantage. Speaking at the G20 Innovation Ministerial held in North Carolina — an event also attended by US Commerce Secretary Howard Lutnick — the OpenAI chief executive said that "a child growing up today will never be smarter than AI," according to KCH FM, citing Infobae. His argument was structural rather than incremental: the gap between human and machine capability will not close, it will widen, and that starting point forces governments to rethink education, work and legislation.

He reinforced the message with a historical comparison aimed squarely at the officials in the room. "I think it would be about as bad an idea to say we're not going to have AI in our country as it was to say we weren't going to have electricity in our country over 100 years ago," he said, per the same report. He paired it with a forecast: "My prediction is that we won't be talking about AI as much as we do now within a decade, because we'll simply expect it everywhere." The implied endpoint is electricity's endpoint — a technology that stops being a topic because it becomes invisible infrastructure.

That analogy is the strongest thing in the speech, and it cuts in more directions than Altman used. Electrification did become universal, but it took decades, enormous public investment, a grid, a safety regulator and a tariff structure — and access remains unequal to this day. "Don't opt out" is a defensible thing to tell a government. "Adopt on the vendor's timetable" is a different proposition, and it is worth noting who is making it: the chief executive of the company that sells the product, addressing the ministers who write the rules. That is not an accusation, it is a disclosure any reader should hold while weighing the argument.

The line about children is the weaker half. Intelligence is not a single scalar with a leaderboard, and we have run this experiment before at smaller scale: pocket calculators have out-calculated every child since the 1970s, and nobody concluded that children should stop learning mathematics. It is also striking how the framing sits against Altman's own recent language elsewhere — he has described AGI as "a very poorly defined term" and "an irrelevant marketing term" in a podcast appearance reported this week. A gap that cannot be closed presumes a measurement everyone agrees on, and the industry does not have one.

Our read: the sentence is fatalism-shaped, and fatalism is the worst possible input to education policy. If machines are permanently better at recall, computation and synthesis, the job of schooling shifts rather than disappears — toward judgment, verification, framing good questions, and orchestrating tools that are confidently wrong a measurable share of the time. That matches what we have argued across our sector work: the teacher who orchestrates AI gains ground; the one who only transmits information loses it. And the long-horizon case Altman is gesturing at is real — tools this capable are the shortest known path to eradicating diseases, extending healthy life and producing genuine abundance. But the transition in between is decided by distribution and governance, not by raw capability. A child who never out-computes a model can still decide what the model is for, who it serves and when to overrule it. That is not a consolation prize. That is the entire job.