The average person's "wealth" -- as measured by life expectancy, access to vaccines, medicines, and advanced procedures, clean food, safe shelter, labor laws (in most countries) plus an enormous variety of entertainment and leisure options -- continues to increase over time, even if wealthy people get the majority of the gains.
Quoting from the article, Elon Musk has apparently said "There will certainly be a lot of job disruption because what’s going to happen is robots will be able to do everything better than us." I don't think that the story is "more of the same for ordinary people, and even worse" if machines can take all routine, productive jobs [1]. If owners don't need ordinary people as employees, then neither do ordinary people need owners. If machines do everything, then the cost (and price) of goods and services falls toward the costs of raw materials and energy. And raw materials and energy themselves fall precipitously in cost if machines can perfectly build more nuclear reactors or solar farms and batteries and mine even low-grade ore deposits without wages or breaks.
In this scenario, ordinary people don't even need to seize the existing assets of rich people to start living well without them. Just have the robots make copies of everything. There are a few edge cases of positional goods where robots can't duplicate things (you can't make more Old Master paintings or Manhattan real estate), but those don't really stand in the way of broad prosperity. If robots do everything routine, then everyone can have good medical care, clean water, plenty to eat, safe and climate controlled housing, abundant electronic entertainment... pretty much everything that the median adult citizen of highly developed countries currently enjoy today.
Keep in mind that I'm not saying that machines will replace human workers in all routine, productive jobs. I'm just thinking about the effects if they can.
[1] Routine, productive jobs would include things like driving trucks, diagnosing and treating leukemia, picking apples, and running experiments as planned by a principal investigator. It wouldn't include things like writing new poetry or coming up with a novel line of research as a principal investigator. But nonetheless, the "routine" jobs include most of today's economic activity, including most that are currently high pay and considered high skill.
Look at how financially powerful a city like Hong King is compared to the issues they’ve struggled with over the last several years.
If we can’t agree on some basic human rights questions, there’s just no way that technology will proliferate in a way that universally benefits everyone.
If you would accept my analogy with a farm, the crop yield per worker may get higher, but instead of working at the field together because we all want that bread at the end of the day, we will be spending most of our time arguing on who's more entitled to get served first.
No one shed a tear when the carriage drivers lost their jobs. Except maybe the carriage drivers and their families.
Or they won't, and we will have a wave a depression, suicide, and populism.
The critical factor is the rate of change.
In the 1880s, there were at least[1] 150,000 horses in New York City. How did the industrial revolution work out for them?
Industrialization was about replacing muscles in niche after niche and muscles are what horses supplied. AI is replacing brains.
1) https://99percentinvisible.org/article/cities-paved-dung-urb...
The idea that new jobs will always arise is not a law of nature, it's only an empirical observation that keeps on being true, until one day it doesn't.
The real problem is more likely to be the transitional turbulence. If self-driving vehicles do wipe out millions of truck driver jobs in the year 2029, but automated doctors aren't ready until 2059, large groups of workers lose income without seeing deflationary relief in the prices of services they need.
[1] The spread of maker machines could happen "legitimately" -- Gates Foundation or a similar organization financing freely shareable re-implementations. It could also happen "illegally" -- a government in a poorer country will just decide to tell international patent/copyright holders to pound sand in favor of their own domestic constitutents. This would be similar to South Africa's treatment of HIV drug patents in the 1990s.