You are not a horse

Treść źródłowa

There’s a popular argument that AI will do to human workers what tractors did to horses. Tractors could do what horses did. Horses became obsolete. AI can do what humans do. Therefore...And every major AI builder seems to agree that humans are next. Musk says AI will “replace all jobs.” Amodei is out there talking all the time about everyone losing their jobs, grounded his framing of AI as “a general labor substitute.” OpenAI investors are out there talking about “80% of all jobs by 2030.” This Substack is reader-supported. To receive new posts and support my work, consider becoming a free or paid subscriber.While these are important people in the space, not some random blogger, they may not exactly be a random sampling of the most knowledgeable people. But it’s certainly not a new fear, and not just an AI thing. Wassily Leontief of input-output fame (more on input-output in a minute) had a few pieces in the early 1980s expressing this worry. If AI really is a perfect substitute for human labor, any cost advantage drives to 100% AI. You don’t need an essay to prove it. But “AI will eventually be a perfect substitute” is doing all the work. But that’s hiding a lot, lots of margins of adjustment and differences, heterogeneity that makes the world the world and not a simple model. How substitutable is AI right now? What would it take for that to rise high enough? What else has to hold? Even the historical example that “tractors could do what horses did, therefore horses became obsolete” sounds like one step, but it’s actually several. And “AI can do what humans do, therefore h…