Thinking Will Become a Hobby: How AI Could Decouple Thought From Paid Work
A recent essay titled 'Thinking Will Become a Hobby' argues that once AI handles most of today's cognitive labor, human thinking will stop being an economic activity and become a leisure pursuit. The case rests on three advantages AI holds over people: better performance, lower cost, and higher availability. It is a bold claim, and the historical parallels are worth testing.
The three-factor argument
The essay's logic is economic rather than philosophical. AI systems are getting better at cognitive tasks, they cost less per task, and they are available at any hour without limit. When all three move in the same direction, an employer has less and less reason to pay a person for the same work.
The point is the combination. Any one advantage alone leaves room for a division of labor, where AI does the cheap part and a human does the rest. Together, they steadily push the human share of paid cognitive work down, and the essay expects thinking itself to be cut loose from earning a living.
What physical labor already taught us
We have seen a version of this before. Machines took over most physical labor during and after the Industrial Revolution, yet people did not stop moving. They kept exercising, but for health and enjoyment instead of survival. Gyms, marathons and recreational sports exist because the work that once demanded strength no longer does.
Games offer another data point. Deep Blue beat Garry Kasparov at chess in 1997, and AlphaGo beat Lee Sedol 4-1 at Go in 2016. People still play both, and professionals now routinely study with AI. The caveat is that games were always leisure. Whether the same holds for law, engineering or research, which people do mainly to earn a living, is the open question.
Who is affected, and how
For software developers and knowledge workers, the pressure lands first on routine analysis, drafting and debugging. The skills likely to hold value longest are framing problems and checking results, though that is an inference from how today's tools are used, not a finding.
For employers, the question shifts from 'can AI do this?' to 'is a person still the cheaper or more accountable option?' Liability, trust and regulation will shape that answer as much as raw capability.
For education systems, the stakes are different. If thinking becomes a hobby, schools may need to justify teaching it for its own sake, not as job preparation. That is speculation on my part, based on how heavily school systems today are tied to career outcomes.
Where this sits next to the 'we need more experts' view
The essay pairs interestingly with an earlier piece here, Terence Tao on why AI-driven math breakthroughs require more mathematicians. Tao argues that humans must keep understanding what AI discovers, which means growing the research community, not shrinking it. Both accept that understanding matters. The difference is who funds it: Tao pictures sustained institutional support for professionals, while this essay pictures thinking as something people do for its own sake.
What would make the prediction wrong
The forecast depends on AI's cost and reliability advantages holding across most cognitive work, not just in demos. It also assumes that societies let labor demand fall without intervening. Regulation, liability rules and human preference for human judgment in high-stakes roles could slow the shift considerably, and the timeline is not something the argument can pin down.
Takeaway
The essay's strongest contribution is the analogy: physical effort lost its economic role but kept its human one. If thinking follows, the real debate is less about whether people will still think and more about who gets paid, who gets time, and how societies share the gains.
Comments
Post a Comment
Please read our comment policy before commenting.