U.S. (and German) automakers were absolutely sure that the Japanese would never be able to touch them. Then Koreans. Now Chinese. Now there are tariffs and more coming to save jobs.
Betting against AI (or increasing automation, really) is a bet against not against robots, but against human ingenuity. Humans are the ones making progress, and we can work with toothpicks as levers. LLM's are our current building blocks, and people are doing crazy things with them.
I've got almost 30 years experience but I'm a bit rusty in e.g. web. But I've used LLM's to build maybe 10 apps that I had no business building, from one-off kids games to learn math, to building a soccer team generator that uses Google's OR tools to optimise across various axes, to spinning up four different test apps with Replit's agent to try multiple approaches to a task I'm working on. All the while skilling up in React and friends.
I don't really have time for those side-quests but LLM's make them possible. Easy, even. The amount of time and energy I'd need pre-LLM's to do this makes this a million miles away from "a waste of time".
And even if LLM's get no better, we're good at finding the parts that work well and using that as leverage. I'm using it to build and check datasets, because it's really good at extraction. I can throw a human in the loop, but in a startup setting this is 80/20 and that's enough. When I need enterprise level code, I brainstorm 10 approaches with it and then take the reins. How is this not valuable?