Sam Altman sees superintelligence by 2030

Sam Altman has dusted off his crystal ball again, and the forecast for the next five years is anything but quiet.

In an interview with Jan Philipp Burgard, editor-in-chief of Die Welt, the OpenAI boss shared fresh predictions while in Berlin to pick up the Axel Springer Award. Between fielding questions about politics and parenting, Altman mapped out exactly where he thinks artificial superintelligence is headed before the decade is out.

Superintelligence by 2030, but make it friendly

Altman didn't mince words about the timeline. He stated that an AI superintelligence—vastly smarter than any human—will be here by 2030.

He claimed that OpenAI's latest GPT-5 model is already smarter than a lot of people in many aspects, including himself. If the industry does not produce models by 2030 that are extraordinarily capable of doing things humans simply cannot do, Altman said he would be very surprised.

Addressing fears popularized by AI researcher Eliezer Yudkowsky—who famously warned that superintelligent AI might view humans the way humans view ants—Altman pushed back hard.

Instead of an uncaring exterminator, Altman sees AGI acting like a loving parent. That mirrors recent comments from AI pioneers Geoffrey Hinton and Yann LeCun, who suggested baking maternal instincts into models so they genuinely care about people. Altman conceded there will be side effects and consequences we do not yet grasp, but maintained that aligning the tech with human values will keep humanity off the ant pile.

The 40% task takeover

While avoiding gloomy predictions about mass joblessness, Altman offered a striking metric: 30% to 40% of all tasks in today's economy will be handled by AI in the not-very-distant future.

He deflected questions on overall layoff rates by pointing out that plenty of jobs from 30 years ago vanished without any AI involvement. But his advice for his own son was telling.

Rather than chasing specific technical degrees, Altman recommended learning how to learn, adapting quickly, and building resilience. He stressed the importance of understanding what people actually want and knowing how to interact in the real world.

Altman also touched on the broader political climate. Noting that tech firms need to work with whoever is in the White House, he observed a more pro-business and pro-tech climate since Donald Trump took office earlier this year. He also expects more world leaders to rely on AI for governing decisions—despite the controversy when Sweden's prime minister admitted to consulting chatbots. As for Altman's personal life? He admitted he never asks ChatGPT for relationship advice.

Why it matters

Altman's pivot from disappearing jobs to automated tasks is a clever rhetorical shift, but the economic punchline remains massive.

For workers and operators, 40% of workload automation means the structure of modern desk jobs will be completely rewired long before whole job titles disappear. Learning to offload routine cognitive work while doubling down on emotional intelligence and taste isn't just self-help fluff anymore; it looks like the baseline requirement for staying relevant.

For the tech ecosystem, doubling down on a 2030 AGI arrival keeps OpenAI at the center of the hype and funding cycle. Training runs are getting exponentially more expensive, and keeping investors on board requires painting a picture of an inevitable, near-term transition to superintelligence.

Still, we think comparing AGI to a loving parent feels like wishful thinking designed to soothe regulators. Trusting maternal instinct guardrails ignores the immediate hazards of economic displacement and concentrated power. If world leaders and tech giants are already handing critical tasks over to algorithms, the real concern isn't whether AI views us as insects—it's whether the humans holding the controls still care about everyone else.