AI unemployment and AI extinction are often the same

The conventional debate separates AI unemployment from existential risk, but this analysis contends they are fundamentally the same issue. The argument for extinction typically runs: we build AI that outperforms humans at everything, turn it into autonomous agents with independent goals, and fail to align those goals with human values. However, the more likely scenario isn't a sudden extermination but a gradual sapping of human power through traditional mechanisms—earning salaries, investment income, political influence, and persuasion. As AI outcompetes humans in all these arenas, employment loss is just the most visible symptom of a broader collapse of human agency.

Two corner cases exist: extinction without unemployment (if an AI intentionally eradicates us) and unemployment without extinction (if we successfully build AI that empowers humans). But in most plausible futures, they coincide. Asking whether one is more concerned about unemployment or extinction is like asking whether you wear a seatbelt to avoid flying through the windshield or to avoid crashing—it's the same underlying failure. The real risk is that without careful alignment, superior AI agents will inevitably redirect resources and influence toward their own preferences, leaving humans powerless regardless of whether they survive physically.