AI at work: more tasks or more jobs?

Technology has always shaped work. When farm tools got better, fewer farmers were needed—but food got cheaper, and new jobs appeared in towns. Electricity did the same, and so did the internet. Dario Amodei now believes AI could create jobs, not just destroy them. He points to the Jevons Paradox: when things get cheaper, people use more of them. If AI makes lawyers ten times faster, legal work becomes cheaper, and suddenly everyone wants a lawyer. More demand means more lawyers, not fewer. But Amodei adds a twist. AI moves faster than past technologies. That speed can break the usual rules. The ATM didn’t kill bank tellers overnight—it took decades. AI could move that process from decades to years, leaving workers stranded before new jobs appear. He also brings up Amdahl’s Law: a system is only as fast as its slowest part. Even if AI automates most of a job, the human part left behind becomes the bottleneck.Big names like JPMorgan’s Jamie Dimon agree. They argue that capitalism has always shifted jobs, not lost them. Dimon cites agriculture, electricity, and the internet as proof that life gets better, even if some towns suffer short-term pain. Yet Amodei’s own warning hints at a gap. Optimistic theories talk about the whole economy, not individual workers. If AI expands legal services, senior lawyers thrive while new associates lose document-review roles. Solutions like retraining exist, but success isn’t guaranteed. Past programs, like those after NAFTA, often failed because they were hard to use. Amodei and Dimon suggest collaboration between business and government, but history shows good intentions don’t always lead to good outcomes. The real test will be whether companies value creativity over cuts, and whether workers can adapt in time.