Sam Altman has executed one of the most striking rhetorical U-turns of the artificial intelligence era, telling an audience in Sydney that the technology he helped unleash on the world will not, after all, trigger the “jobs apocalypse” that doomsayers, including, until recently, himself, have spent the past three years forecasting.

Speaking at a Commonwealth Bank of Australia conference, the OpenAI chief executive conceded that his predictions about how quickly ChatGPT would hollow out the white-collar workforce had been “pretty wrong”. It is an unusually candid admission from a Silicon Valley founder more accustomed to selling the future than apologising for the one he predicted.

“I’m delighted to be wrong about this,” Altman said. “I thought there would have been more impact on entry-level white-collar jobs being eliminated by now than has actually happened. I don’t think we’re going to have the kind of jobs apocalypse that some of the companies in our space advocate or talk about.”

For Britain’s small and medium-sized businesses, many of whom have been wrestling with whether to bet the farm on AI tooling, or to hold their nerve and keep hiring graduates, the comments will land somewhere between reassuring and infuriating, depending on how much capital they have already redirected into the technology on the strength of earlier warnings.

The human factor altman says he underestimated

Altman’s explanation for his change of heart is, on the face of it, refreshingly mundane. The chief executive recounted experimenting with using AI to handle his own Slack messages, only to find the exercise served as “an amazing example to me of we really do care about people”. The interactions he most values, he said, were “not something that I can imagine myself outsourcing to an AI anytime soon”.

It is a far cry from his earlier suggestion that entire categories of work, customer support roles in particular, would be wiped from the economy. The pivot is consistent with a 19th-century paradox economists have long pointed to: the more productive machines become, the more valuable distinctly human attributes, judgement, empathy, accountability, seem to grow. As Business Matters has previously reported, a sizeable cohort of technologists has consistently argued that AI is a tailwind for skilled labour rather than the destroyer-of-worlds presented in some quarters.

Not everyone in silicon valley has got the memo

The trouble for Altman is that his peers do not appear to share his newfound optimism. Dario Amodei, chief executive of rival lab Anthropic, warned only last year that AI could wipe out half of all entry-level white-collar jobs and drive unemployment to between 10 and 20 per cent within five years, a forecast he set out in a widely discussed interview with Fortune that has aged uncomfortably well for those tracking corporate redundancy announcements.

The list of household names cutting headcount and citing AI as a reason has lengthened considerably in recent weeks. Standard Chartered confirmed plans to eliminate roughly 7,800 back-office positions, about 15 per cent of the affected workforce, by the end of the decade, with chief executive Bill Winters describing the cuts as a replacement of “lower-value human capital” with technology. The phrase prompted such a backlash that Winters was forced to send a clarifying memo to staff within days.

Meta, meanwhile, has begun another sweep of layoffs, shedding 8,000 roles, roughly a tenth of its workforce, while Amazon, Microsoft and Jack Dorsey’s Block have all announced significant cuts this year. According to industry tracker Layoffs.fyi, more than 140 technology companies have collectively let go of in excess of 111,000 staff since January.

The UK picture: graduate schemes feeling the pinch

For the SME owner-managers who form the backbone of Business Matters’ readership, the more telling indicator is what is happening at the entry-level end of the talent pipeline in the UK itself. Graduate vacancies, particularly in professional services, have softened markedly, and entry-level job postings have fallen by close to a third since ChatGPT’s launch, with retail, IT and finance bearing the brunt. The big four accountancy firms have trimmed graduate intakes as junior tasks are absorbed by automation.

Whether that constitutes an apocalypse or simply a structural reset is a matter of vocabulary. What is harder to dispute is that the bottom rung of the corporate ladder is shorter than it was three years ago, and one in six UK employers now expects to make further AI-related cuts within the next twelve months, according to recent industry surveys.

What the Goldman numbers actually say

The most reliable view of the labour market sits, as so often, with the bean-counters at Goldman Sachs. The bank’s economists estimate that AI has trimmed monthly US payroll growth by roughly 16,000 jobs over the past year, and nudged the unemployment rate up by around 0.1 percentage points. Roles most exposed, telephone operators, insurance claims clerks, financial administrators, are seeing measurable contractions in headcount. Conversely, jobs where AI augments rather than replaces human effort have added around 9,000 positions over the same period.

In aggregate, these are not the numbers of an apocalypse. They are the numbers of a slow, asymmetric shake-out, the kind of structural change that hits the unprepared hardest and rewards those who adapt early.

What this means for British SMEs

For the owner of a 50-person accountancy practice in Manchester or a marketing agency in Bristol, Altman’s volte-face is less an instruction to relax than an invitation to think more clearly. The replacement-by-AI thesis is plainly happening at the largest, most process-heavy employers, banks, the Big Four, hyperscale technology firms. It is happening considerably less, and considerably more selectively, in the small and medium-sized businesses that employ around three-fifths of the UK’s private sector workforce.

The lesson, if there is one, is that the question is no longer whether AI replaces people. It is which people, doing what tasks, in which businesses. Altman, for now, thinks his earlier prediction was the wrong one. SME leaders watching their hiring pipelines, redundancy budgets and graduate schemes would be wise to draw their own conclusions, and to act on them before the next U-turn.

Read more:
Altman backs away from AI ‘jobs apocalypse’ warnings as OpenAI chief admits he was ‘pretty wrong’