Knowledge workers are losing faith in their careers as AI closes in

Knowledge workers are losing faith in their careers as AI closes in

Aaron Horwath opens with a scene from a train commute: across from him sat a man in his early 30s, slacks and dress shirt, AirPods in, who spent over half an hour on a call explaining EBITDAs, margin expansion opportunities, cost structures and ARR in monotone detail. When the train stopped, the man dug out two knitting needles and a mound of pink yarn and explained he was making a winter hat for a niece, a project born, as he put it, from a desire to do something. Horwath writes that he noticed a glint of pride and excitement in the man's eyes that had been absent during the business call.

Horwath says he hears versions of this more and more among his knowledge-worker peers: people picking up pottery, painting or crochet, sharing dreams of disappearing, living on a farm, or going off the grid. He frames these not as idle daydreams but as symptoms of a rising, indescribable melancholy and a doubt that knowledge work has ever had a point, a doubt he says now reaches even highly paid executives and senior professionals who are otherwise the most insulated from AI's near-term economic impact.

To explain the mechanism, Horwath turns to Derek Thompson's 2019 Atlantic essay on 'American Workism,' quoting Thompson's description of it as 'emotional, even spiritual. The best-educated and highest-earning Americans, who can have whatever they want, have chosen the office for the same reason that devout Christians attend church on Sundays: it's where they feel most themselves.' Horwath contrasts true vocations, teaching, nursing, firefighting, skilled trades, with the finance, consulting and technology jobs most graduates now take, which he says are demanding and competitive but often lack any altruistic upside. He cites David Graeber's 'Bullshit Jobs' for its catalog of pointless corporate labor: slide decks for projects that never launch, feature debates nobody asked for, 'resting and vesting,' and promotion-driven development. Workism, in Horwath's reading, is what lets people do this work without losing their minds: a substitute faith that requires believers to keep showing up.

He then argues AI threatens that faith by adding a further layer of abstraction to work that was already, in Graeber's framing, structured like Russian nesting dolls of roles supporting roles. Horwath notes it is now common, himself included, among people who integrate AI tools into organizations, to describe the future of work as humans managing armies of AI agents. But he says employees under pressure from leadership are already going well beyond delegating grunt work, using AI agents to formulate complete business strategies, generate full marketing campaigns, build entire websites and draft strategy for whole divisions.

Horwath closes the argument with Guy Debord's 'The Society of the Spectacle,' quoting its opening line that 'life is presented as an immense accumulation of spectacles. Everything that was directly lived has receded into a representation,' and Thesis 12, that spectacle's sole message is 'what appears is good; what is good appears.' He casts workism as a microcosm of that spectacle, where appearing busy and important substitutes for actually being so, and argues that adding AI moves workers even further from the value of their own labor: instead of writing the pitch, he writes the query that tells an AI to write it, then checks the result. Even when that underlying work felt unsatisfying before, he writes, it was still ours.

Key facts

  • Aaron Horwath, director of AI operations at a creative technology company, writes the essay for Noema.
  • Opening anecdote: a commuter in his early 30s spent over half an hour on a business call about EBITDAs, margin expansion and ARR, then knitted a hat for his niece once the train stopped.
  • The essay builds on Derek Thompson's 2019 Atlantic concept of 'American Workism,' the idea that professionals seek meaning from work the way earlier generations sought it from religion.
  • It cites David Graeber's 'Bullshit Jobs' (canceled-project slide decks, 'resting and vesting') and Guy Debord's 'The Society of the Spectacle' to argue white-collar work already substituted appearance for substance before AI arrived.
  • Horwath argues AI agents add a further layer of abstraction, since employees increasingly direct AI to write pitches, strategies and newsletters rather than doing that work themselves, a shift he describes as workers becoming 'managers of armies of AI agents.'

Why it matters

This is a mood piece rather than a data piece, but it names something concrete: a discourse tying AI's move into cognitive labor to a specifically existential, not just economic, crisis. Horwath's argument is that workism, the substitution of career for identity that Derek Thompson diagnosed in 2019, depended on workers still being the ones who did the work. AI agents doing the actual drafting, strategizing and reporting removes that last psychological anchor, which is why the anxiety he describes shows up even among senior, economically insulated professionals rather than only those facing layoffs.

Who it affects

Knowledge workers in finance, consulting and technology, the demographic Thompson's original workism argument targeted, especially highly paid executives and senior professionals with deep institutional knowledge, whom Horwath says are usually the most insulated from downturns yet report the same career anxiety. It also implicates people like Horwath himself, who as a director of AI operations at a creative technology company sits on the implementing side of the shift he describes.

How to use it

As an essay rather than a tool, its use is diagnostic. Horwath's frame, that workism depended on people doing the work rather than merely directing it, offers a lens for recognizing why AI delegation can feel unsettling even when it removes drudgery. Anyone deciding how much of their own work to hand to AI agents may find the Graeber and Debord references useful for naming the discomfort, separating tasks that were genuinely bullshit to begin with from ones whose meaning depended on personal execution.

How solid is it

The piece is a first-person opinion essay, not a reported or data-driven feature. Its central claim, that disillusionment is spreading among knowledge workers, rests on Horwath's own train anecdote and unnamed conversations with his 'Knowledge Worker peers'; no survey, poll or study is cited to quantify how widespread the trend is, and no statistic on job losses, layoffs or AI adoption is given. The essay's authority instead comes from its citations, Derek Thompson's 2019 Atlantic piece on workism, David Graeber's 'Bullshit Jobs' and Guy Debord's 'The Society of the Spectacle,' which are accurately attributed and quoted.

Risks and caveats

Because the evidence is anecdotal, one overheard conversation and informal peer talk, the claim of a spreading, era-defining existential crisis should be read as one observer's pattern-matching rather than a measured trend. The commuter's job and employer are never confirmed beyond his talk of EBITDAs and ARR, and Horwath's own employer goes unnamed too, so the specifics that would let a reader check the framing against reality are largely absent.

“Workism is emotional, even spiritual. The best-educated and highest-earning Americans, who can have whatever they want, have chosen the office for the same reason that devout Christians attend church on Sundays: it's where they feel most themselves.”

— Derek Thompson, quoted in the essay