OpenAI chief executive Sam Altman said on 2 November that at the start of this year “AI was mostly not as smart as our smartest scientist and now it’s mostly smarter than them,” while calling the comparison “very approximate because it’s spiky and it’s not true in all ways.” He spoke at the opening of ADIPEC, the Abu Dhabi energy conference.
He made the same before-and-after claim about office work: “AI was mostly not able to do a lot of white collar jobs and now it mostly can do a lot of parts of that.” In US payroll data, information-sector jobs are down 10.5% since late 2022, about 325,000 jobs, while professional and technical services added about 163,000, by our count of Bureau of Labor Statistics figures we reported earlier today.
Watch · Sam Altman at the ADIPEC 2026 opening
The ADIPEC 2026 opening session with Sam Altman’s keynote and Q&A, as posted by DRM News on YouTube on 2 November. Timestamps in this piece refer to this video.
The key insight: Sam Altman said that during 2026 AI went from mostly below to mostly above “our smartest scientist,” calling it very approximate; US payroll data show information jobs down 10.5% since late 2022 while professional and technical services added about 163,000, by our count of BLS figures.
Altman gave both comparisons in the Q&A with The National’s editor-in-chief Mina Al-Oraibi, at 11:56 to 12:13 in YouTube’s automatic captions of the session video, in answer to a question that included people’s nervousness about AI’s impact.
In his keynote he described AI as “a sort of virtual colleague or teammate,” with people getting “much more work done with fewer people and and much more leverage,” as the captions record it. He also said: “in the last few months we have seen AI begin to solve math problems that people have been unable to solve.”
Payroll jobs in three AI-heavy US sectors, quarterly average of monthly seasonally adjusted counts, 2022 Q4 = 100, to 2026 Q2. Source: BLS Current Employment Statistics (latest months preliminary). Index values and job changes are our arithmetic, as reported on 2 November.
Not Dreaming Big Enough
Asked about the ADNOC agreement, Altman said: “we’re all not dreaming big enough,” adding: “because we’re all kind of used to last year’s models.” He said that “we’ve got to go find the biggest problems to take on that we can.”
He said OpenAI has started describing the moment differently from an industrial revolution: “it feels much more like a new renaissance.” On jobs, he said: “We’ll find ways to adjust our jobs.”
What the Payroll Data Shows
Our count of BLS Current Employment Statistics compares quarterly averages for 2022 Q4 and 2026 Q2. Information fell about 10.5%, finance and insurance about 0.6% (about 37,000 jobs), and professional, scientific and technical services rose about 1.5%. The three together lost about 199,000 jobs, or 1.0%.
Those payroll figures show where jobs moved, not why they moved, and they do not test Altman’s claim. BEP Research’s chart, which we covered this morning, says real output in the same three sectors rose 19.8% over the period.
As we read it, Altman made two separate before-and-after claims in one answer, one about the top of the skill range (the smartest scientists) and one about the breadth of office work (“a lot of parts” of white-collar jobs), and gave a measure for neither.
As we read the BLS series, the payroll split runs inside the three sectors rather than across them: the information sector shrank while professional and technical services grew, which says where jobs moved but not why.
Sam Altman, at the ADIPEC 2026 opening in Abu Dhabi (YouTube automatic captions)
“we’re all not dreaming big enough”
Three Implications
FOR EMPLOYERS Altman said AI was mostly not able to do a lot of white-collar jobs at the start of the year and now “mostly can do a lot of parts of that,” while saying “We’ll find ways to adjust our jobs.”
FOR WORKERS IN INFORMATION JOBS By our count of BLS data, information-sector payroll jobs fell about 10.5%, or about 325,000, between 2022 Q4 and 2026 Q2; the data do not show the cause.
FOR AI BUYERS Altman said “we’re all kind of used to last year’s models” and that “we’ve got to go find the biggest problems to take on that we can.”
The framework argues that agent services are disrupting work because “human execution was the assumption underneath all previous layers.”
As we read it, Altman’s claim that AI “mostly can do a lot of parts” of white-collar jobs is a lab chief’s version of the same premise. The framework is the owner’s argument and does not discuss Altman’s remarks or the payroll data.
Altman gave no measure for either comparison, and did not say which scientists or which jobs he meant. Nothing in the BLS data links the information-sector decline to AI. The quotes come from YouTube’s automatic captions of a 21-minute video of the session posted by DRM News on 2 November, not from an official transcript; we corrected only the spelling of names.
This Week In The AI Supercycle
Anthropic’s Reality Check, The Capital Squeeze, & The New World Order
This week’s issue: for two years the AI build-out was priced as a growthstory. This week it started being priced as a credit story.
Altman said that within this year AI went from mostly below to mostly above the smartest scientists, and from mostly unable to mostly able to do a lot of parts of white-collar jobs, calling it very approximate. US payroll data show information jobs down 10.5% since late 2022 while professional and technical services grew, by our count.
A note on sourcing. This piece relies on YouTube’s automatic captions of the ADIPEC 2026 opening session video (posted by DRM News on 2 November 2026), read in full, and BLS Current Employment Statistics payroll series. Timestamps refer to that video. Captions are machine-generated; we corrected only the spelling of names. Quarterly averages, index values and job changes are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
