That’s actually not true. The Baumol cost disease means that people who are not actually more productive in their current roles (eg orchestral clarinetists) still make more money because of productivity growth elsewhere. You couldn’t possibly pay an orchestral clarinetist as much today as you did in 1800 because they’d just quit and do literally anything else for a living. You have to pay them more.

On a scale from 1800 to now, we are all better off. But if you look at the past 20 years in the us, we are clearly not benefiting all of society. Arguing that the avg clarinetist is slightly better off today compared to 1970 doesn't account for the fact that many many other people haven't had an increase, but they've had a decrease in their standard of living.

I find this claim rather incredible, if only because of the set of things that hadn’t even been invented in 1970, which nonetheless everybody seems to have today.

> Productivity increases due to automation go to the owners of capital.

That is only true if there is a monopoly on automation, which there generally isn't. In which case the productivity increases go into lower prices, which result in greater demand for whatever is still scarce.

That can be capital owners (e.g. land), but it can also be skilled labor or literally anything else which is in demand and people choose to spend the money on that they had saved from the lower prices driven by automation.

Whatever the reasoning, data shows wealth and income gaps increasing, and the ones with capital are the on the winning side.

That has nothing to do with automation and everything to do with low interest rates. When interest rates are low, people borrow more money and bid up the prices of capital assets.