Indeed. Robots aren't intrinsically different from, say, combines. That doesn't mean that it isn't worth looking at disruptions caused by productivity growth, but I think we'd be better off looking at what kinds of disruptions were seen in the past. This not only gets us away from the robots narrative (which seems to muddle things a bit as it feeds into present day hype), but it also allows us to look at the impact of much higher productivity growth (current productivity growth is fairly low).

Yes, it's not liquid (mainly due to specialization, current robots designed to do something in one factory aren't that usable doing something a bit different elsewhere), but it's still a long term investment.

If you pay $1M in salaries, that won't help you next year at all; if you buy robots for $1M, you'd expect at least half of them to still do useful work for you after quite a few years.