The first person to make a useful product with it is going to dominate. I think a functional robot is probably going to be the system that really gets it.

Also web design is (for good reason) going to be one of the first to go. But as far as actually disrupting the economy as a whole it'll be YEARS before that happens.

But to be honest, besides replacing indian call centers, chat bots really aren't that "useful" yet

So the race to build AI is basically just going to be a money suck until something clicks. Again. probably a useful robot.