Sam Altman has spent three years telling everyone that the disruption was almost here. On David Senra’s podcast, the OpenAI chief executive finally said the quiet part out loud: the clock he keeps setting has been wrong. “I thought when we got to GPT-4, which was back in 2023, that very quickly after that there was going to be much more disruption, software businesses up for grabs right away, than it turned out to be,” he said.What he blames is not the technology. It is us. “I think I was wrong about a few things, but one in terms of the speed: the economy just has so much inertia,” Altman said, pointing out that people keep buying from the same vendors and using the same tools in the same way. He framed that stubbornness as good news, saying it will make the transition “smoother and slower” and that he is grateful for it. Then came the line that will follow him around: “But it means we’ve all been too ambitious on timelines. Even with this incredible technology, society and the economy will adapt more slowly.”
Sam Altman’s GPT-4 prediction did not survive contact with the real economy
The specificity here is what stings. In his January 2025 blog post Reflections, Altman wrote that OpenAI was “now confident we know how to build AGI as we have traditionally understood it,” and predicted that 2025 “may see the first AI agents ‘join the workforce’ and materially change the output of companies.” By June 2025, in The Gentle Singularity, he had a year-by-year schedule: agents doing real cognitive work in 2025, systems producing novel insights in 2026, robots handling physical tasks in 2027. On the podcast, the man who wrote that is describing an economy that barely noticed.
The AGI timeline has quietly slipped from 2025 to 2030
Altman’s own forecasts have been sliding for a while. Earlier gestures at 2023 and 2025 gave way to 2030 by late 2025, a revision that landed while OpenAI’s finances were being picked apart. The company was valued near $500 billion despite reported losses of $5 billion on $3.7 billion in revenue, and it does not expect a first profit until 2029. There is also the awkward contractual detail: the Microsoft agreement reportedly ties AGI to a $100 billion profit figure, which makes any timeline announcement partly a financial statement.
Rival AI bosses have been making this argument about AGI for years
None of this is news to Altman’s competitors. Microsoft AI chief Mustafa Suleyman said flatly that the uncertainty around AGI is so high that categorical declarations feel ungrounded. Demis Hassabis has put human-level reasoning at least a decade out. Yann LeCun has insisted it will not arrive as a single event. Andrew Ng told people to be sceptical of any company claiming AGI is imminent. Dario Amodei was the outlier, forecasting a “country of geniuses in a data center” by 2026.So the AI race now has its loudest accelerationist arguing for patience. The technology, by his account, still works. The world it was supposed to upend simply had other plans, and Altman says he is thankful for the delay.