OpenAI CEO’s DevDay comments sparked backlash as mounting evidence shows AI already replacing human roles across multiple industries.
Sam Altman triggered controversy at OpenAI’s DevDay conference earlier this month with remarks suggesting many jobs eliminated by AI might not constitute “real work” anyway. Speaking with AI newsletter founder Rowan Cheung, the OpenAI CEO used a thought experiment about historical perspective to frame his argument.
Altman proposed that a farmer from 50 years ago would likely look at modern knowledge work and question its legitimacy. “The thing about that farmer is they very likely would look at what you do and I do and say, ‘that’s not real work,'” Altman stated. He contrasted farming, which produces tangible necessities people need, with contemporary white-collar employment.
The comments were immediately clipped and circulated across social media, drawing harsh criticism. Many labeled Altman’s perspective as callous or dystopian, particularly given his position leading a company whose technology directly threatens knowledge worker employment.
Historical Context for ‘Bullshit Jobs’ Theory
Altman isn’t breaking new ground with this argument. The late anthropologist David Graeber published “On the Phenomenon of Bullshit Jobs” a decade ago, arguing that many workers secretly believe their positions lack meaningful purpose. The essay went viral before expanding into a bestselling 2018 book.
Graeber claimed entire economic sectors exist primarily for box-ticking bureaucracy without generating genuine social value. His thesis resonated widely, cited by everyone from frustrated office workers to policy think tanks examining labor market efficiency.
However, academic research hasn’t substantiated these sweeping claims. A 2021 study using the European Social Survey found only about five percent of respondents considered their jobs useless. A comparable U.S. study put that figure closer to twenty percent.
Critically, these researchers concluded that feelings of pointlessness stemmed more from poor management and toxic work culture than from the inherent nature of the roles themselves. When micromanagement and broken workflows dominate, even valuable work can feel meaningless. This suggests organizational dysfunction rather than fundamental job illegitimacy.
Where Altman’s Argument Holds Merit
Despite the overreach, Altman’s comments contain a kernel of truth about task composition in modern work. Most jobs aren’t fake, but many have accumulated layers of automatable administrative overhead: compliance checklists nobody scrutinizes, reports filed but never read, emails summarizing meetings that could have been brief asynchronous messages.
This is precisely where large language models excel. When Altman discusses AI eliminating work, he likely refers to these repetitive, low-value tasks rather than entire professions. LLMs already handle routine correspondence, generate standardized documentation, and process information retrieval more efficiently than humans in many contexts.
The distinction matters enormously. Eliminating administrative busywork differs fundamentally from replacing skilled judgment, creative problem-solving, or relationship-building components of knowledge work. Yet Altman’s framing blurred these lines, suggesting entire roles might disappear rather than emphasizing task-level automation.
Mounting Evidence of AI Job Displacement
Altman’s comments arrive amid growing documentation of AI replacing human workers. Multiple industries report workforce reductions as companies deploy automation tools. Customer service, content moderation, basic coding, and data entry positions face particular pressure.
The displacement isn’t purely hypothetical. Companies increasingly announce headcount reductions explicitly attributed to AI adoption. These aren’t speculative future scenarios but present-day business decisions reshaping labor markets in real time.
However, the narrative that AI eliminates only “non-real” work provides convenient justification for these layoffs while obscuring more complex dynamics. Many displaced workers performed legitimate functions that companies simply decided could be automated cheaply enough to justify the quality tradeoffs.
The Optics Problem for AI Leadership
For someone in Altman’s position, dismissing threatened jobs as potentially illegitimate carries particular weight. OpenAI’s technology directly enables much of this displacement. Framing job losses as revealing work that wasn’t valuable anyway shifts blame onto workers rather than acknowledging the economic disruption his company facilitates.
The farmer comparison also reveals class assumptions about what constitutes “real” work. Physical labor producing tangible goods gets validated while knowledge work faces skepticism, despite both being economically necessary in modern societies. This hierarchy of work legitimacy doesn’t withstand serious scrutiny but persists in public discourse.
Altman’s broader point about workflow inefficiency deserves attention even if his delivery failed. The challenge is improving work quality and removing genuine busywork without using that analysis as cover for mass workforce reductions driven primarily by cost optimization rather than productivity gains.