> and just assumed that there must be some way to automate and streamline almost everything, as long as the demand is there
Scaling a solution to the travelling salesman problem would solve many other problems. The demand is definitely there. But many people believe it cannot be done.
That's a fundamentally different problem. "Scaling" it requires figuring out how to solve a different concrete problem every time. Scaling e.g. battery production requires figuring out how to solve the same real-world problem very efficiently.
No, you could (possibly) solve TSP once by providing an algorithm that scales well just like inventing an industrial process. My point was that it looks very much like some problems are fundamentally difficult and have no scalable solutions.
Scaling a solution to the travelling salesman problem would solve many other problems. The demand is definitely there. But many people believe it cannot be done.