Cognitive Mercantilism: AI panic resembles fears of early capitalism

Discussion overview

The thread mostly rejected the article’s reassuring historical analogy. Commenters argued that earlier automation shifted people from farm labor into intellectual work because both physical strength and cognition were human assets; AI threatens to automate the latter too, leaving no obvious next capability through which most people can earn bargaining power. The concern was not merely unemployment, but ownership: frontier models, training data, and data centers look far easier to concentrate than land, tools, or human know-how, letting a small group rent out intelligence while everyone else becomes dependent on access.

Several readers also challenged the article’s rosy account of capitalism and trade, arguing that past prosperity was inseparable from colonial extraction, labor conflict, and unequal distribution. On that view, invoking previous transitions says little about whether AI gains will reach ordinary people without deliberate political and economic reform. A smaller group pushed back: present AI remains narrow, particularly weak where subjective human appeal and social judgment matter, and every deployment still needs people to choose goals, supply context, absorb risk, and own the consequences. They see AI as relocating work rather than abolishing human relevance. Even those less alarmed accepted that a world of broad automation would require institutions to distribute security and agency, not just new personal hobbies.

Points of disagreement

The sharp dispute is whether AI will repeat prior technological transitions by moving human value into new work, or end that pattern by automating the last broadly marketable human advantage. Optimists argue that AI is still bounded by human context, judgment, and demand, while critics argue that centralized ownership of machine intelligence can erase labor’s leverage before any replacement role emerges.