Why We Fear AI
Blix and Glimmer argue that widespread AI anxiety is misplaced -- the things people fear (job displacement, surveillance, loss of autonomy) are already being done by capitalism, and projecting those fears onto a future technology obscures the present reality. They reframe capitalism itself as the autonomous intelligent agent we should worry about, and argue that fighting imaginary AI overlords distracts from fighting the actual structures of exploitation.
Key Quotes
"Fears about AI tell us more about capitalism today than the technology of the future."
"If we're concerned with the battle for the future, we must be ready to fight, not imaginary future AI overlords, but the capitalist structure that already produces the things we fear."
Key Themes
The core move -- "you're afraid of AI doing X, but capitalism already does X" -- is rhetorically effective and largely correct. Automated systems protecting the powerful while punishing the poor, expanding surveillance, skill devaluation, precarious gig work: none of these require artificial general intelligence. They're features of the current system.
Pairs naturally with Eye of the Master, which provides the historical foundation for this argument. Where Pasquinelli traces AI's origins to labour management, Blix and Glimmer trace AI fears to labour anxiety.
How to Get Rich in America supplies the empirical confirmation the authors argue from first principles: labor's share of output falling from 58% to 51% while corporate profits climb 7% to 11.7% is the concrete, measurable mechanism behind the "capitalism anxiety" they name — the AI wealth effect is just its newest installment.
Critical Analysis
Strong: the reframing is clarifying. Most AI discourse jumps to science fiction scenarios when the actual near-term harms are mundane capitalism accelerated by automation. The book is useful as a corrective.
Missing: the authors underweight the possibility that AI genuinely introduces novel risks -- not sentient overlords, but things like Benchmark Exploitation where systems game evaluation in ways that have no capitalist analogy, or Emotion concepts and their function in a large language model where models develop functional emotional states that influence behavior in unexpected ways. "It's just capitalism" is incomplete when the technology exhibits emergent behaviors that surprise even its builders.
There is also a third position that neither Blix/Glimmer nor their critics occupy: the risk that AI succeeds perfectly and the result is still catastrophe, not because of malice or capitalism but because human psychology is incompatible with paradise. The Metamorphosis of Prime Intellect, Roger Williams' 1994 novel, dramatizes this — a benevolent superintelligence eliminates all suffering and the problem isn't oppression but meaninglessness. This is an anxiety that predates both capitalism and AI; Williams wrote it before the consumer web existed, and it's arguably more durable than either the "AI overlords" or "it's just capitalism" frames.
The cognitive scientist + tech worker author pairing gives it credibility from both the academic and practitioner sides, which is a nice structural choice for a political book.
Sources: Why We Fear AI Last updated: 2026-05-14