Why Common Arguments About AI Replacing Jobs Fall Short

Summary

This opinion essay, presented as a transcript from a Lex Fridman Podcast interview with Sam Altman, challenges the argument that AI cannot replace human work simply because chess engines did not eliminate chess. Humans continue to play and watch chess because the activity has social, competitive, and entertainment value, while AI plays the game in a fundamentally different way. The author argues that programming is less protected by this analogy: it is a creative problem-solving activity, but if an AI could solve programming tasks instantly and at scale, companies would have little reason to retain large teams merely to produce similar code. Human-made art may retain emotional and cultural value even when AI automates some production, whereas people may have no reason to prefer human-written software if its behavior is equivalent. The essay also rejects the comparison with calculators, noting that calculators perform arithmetic while mathematicians work with axioms, conjectures, and theorems; a more advanced theorem-generating AI could therefore threaten mathematical work more directly. The author speculates that programming jobs could disappear roughly from well-documented, less complex web development toward more specialized and theoretical fields as AI improves, although a genuinely superintelligent general AI might affect them all at once. The main fear expressed is not an extinction scenario but the possibility that human intelligence becomes unnecessary, removing the problem-solving activities the author values. The author is uncertain whether that future will occur and says releasing open-source code is the most practical way to leave some influence on a possible future superintelligence.