The grieving economy: what the AI boom is doing to the workforce

In previous editions of The Counterbalance, we have tracked how big tech and concentrated private capital are setting the direction of travel for large swathes of the economy ranging from manufacturing to insurance for household robots.

In either example, which can be read here and here, we note how rampant speculation on AI is generating products and services that would have little reason to exist were it not for the capital that preceded them.

But unlike either of these accounts, where the harms are still partly prospective, the damage AI is dealing to the workforce is already clear and obvious. Jack Maguire, who has spent over a decade building advertising programmes across some of the largest tech companies including Meta and TikTok, recently wrote an article looking at the collective psychological toll behind AI.

In his piece, he documents how a specific form of grief is spreading through the online communities, including Reddit threads that have accumulated tens of thousands of individual engagements.

“AI-driven displacement is producing a distinct emotional category that most closely resembles grief, distinct from ordinary fear, anxiety or burnout. That grief is structurally suppressed, because layoffs are framed as routine business decisions that leave no socially sanctioned room for mourning,” Maguire said. “And the standard grief model itself breaks down in the AI case, in a specific way that makes recovery harder than it was in previous industrial transitions.”

Maguire draws on a Sharma (2025 et al) study in the International Journal of Qualitative Studies on Health and Well-being which found that participants experienced AI-related job displacement as “the symbolic loss of professional identity, autonomy, and future prospects.” On Reddit, two individuals working in data science and analytics posting had not lost their jobs but had described grieving a sense of meaning to their work.

“The data communities have spent the past year documenting a bifurcation of the [general] data scientist, squeezed from above by machine-learning engineers and from below by analysts equipped by large language models,” Maguire also said.

The core of Maguire’s argument is that AI displacement reaches past income and job security, and touches personal identity in a way which previous industrial or technological transitions did not. As the Sharma study found, workers have experienced “the symbolic loss of professional identity, autonomy, and future prospects.”

Economists Brett Hemenway Falk and Gerry Tsoukalas build on the structural factors at play in a model which they call the “AI Layoff Trap”. Its central finding is that competitive markets create a demand that drives firms to automate beyond what is collectively within our public interest.

A clear example of this is when a company replaces workers with AI, or at least tasks with AI, in order to save costs. Yet, the demand it destroys (because displaced workers are also consumers) spreads across the whole economy. The authors describe this as a trap: “an automation arms race that only intensifies as AI improves, that leaves workers and firm owners alike worse off, and that no market force can break.”

Dario Amodei, Anthropic CEO and the man behind one of the most widely used AI models (Claude) himself warned that AI displacement would be faster, broader than previous technological revolutions and “unusually painful,” particularly for those with less market power than the monopolists. “The model points, perhaps counter-intuitively, to where the problem is most severe: not dominant technology firms but fragmented industries,” said Falk and Tsoukalas.

In other words, the largest technology firms, the very companies behind the AI boom in the first place, are essentially insulated from the trap described by the authors by virtue of their market dominance.

The harms thus are carried by the smaller actors in competitive markets that cannot afford not to deploy AI technology because they must keep pace with rivals. Falk and Tsoukalas identify customer support sectors where thousands of firms replace agents with AI, and software services where tools enable companies to trim a multi-person team to a singular engineer, as two example sectors which bear the cost the most.

“Each firm reaps the full savings of replacing its own workers yet bears only a sliver of the demand it destroys; the rest lands on rivals. No firm can afford to be the one that holds back,” they added.

Both Maguire, and Falk and Tsoukalas arrive at the same destination, albeit from different directions or perspectives. The damage done by the burgeoning AI sector is both inward (to the individual who can feel the meaning draining out of their relationship with work) and simultaneously on a market structure level.

Together, these two perspectives account for a system in which the people least responsible for (and least able to control) the AI boom are absorbing most of its costs, whereas the big tech giants that have orchestrated it retain their power and dominance in the political economy and public life.

Falk and Tsoukalas propose a tax which would impose a levy on companies each time they replace a human with AI. This tax, a “Pigouvian automation tax”, acts as a means of disincentivising activities that generate harms like unintended costs on third parties. It would, at least in theory, prompt companies to reconsider the social cost of their actions and discourage otherwise reckless behaviours.

Balanced Economy Project would go a step further and argue that policymakers must target the root of the power in the AI market economy to promote systemic transformation. Constraining concentrated economic power at the structural level requires steps like breaking up the concentration at the AI infrastructure level, ending single-vendor dependency in public sector AI procurement, and embedding competition principles in AI industrial policy.

Dear reader, please help us make The Counterbalance better by filling out a quick reader survey. We would love to hear form you:

https://thecounterbalance.substack.com/survey/7403461

The Counterbalance is published every Thursday. Please send any thoughts and feedback to scott@balancedeconomy.org.