I already automated the uninteresting problems, so the ones that I work on are the interesting ones that remain. Once an AI can do what I currently do better than I can, the next-most-valuable thing I can be hired for is repetitive manual labour, where I'll be competing on a cost and speed basis with servo motors.

> Once an AI can do what I currently do better than I can, the next-most-valuable thing I can be hired for is repetitive manual labour

You move one layer up or down the stack. Build things that support AI, or orchestrate it.

One thing humans have and AI models don't is private personal experience. Your perspective can still be useful even if AI can do many things on its own.

Another thing humans have and AI don't - responsibility, you can't do anything to a bad AI, you can punish a person, so we can trust people more where there is a big stake.

> Build things that support AI, or orchestrate it.

Is that your job? What will you do when AI does it better, faster, and cheaper than you can do it?