The article assumes that undeveloped areas will suffer from these changes, because the old path to development will not be available to them.
But new paths to development will become available, from these same technologies.
Increasingly, undeveloped areas will find it easier, cheaper and faster to just make their own stuff locally. Robots will get better and cheaper, just like cell phones did - everyone will get them, and sooner than folks thought ahead of time.
Artificial intelligence will increasingly replace the need for an educated workforce, potentially REDUCING the comparative advantage of more developed societies. Those who invent new things could make huge fortunes quickly, but the bulk of the human population will quickly be able to get their own copy at little incremental cost (a software update).
If undeveloped countries don’t have the capital to buy their own production equipment, just leave the door open, and foreign companies will set up shop to produce locally, as it becomes more economical - prices will still drop for their consumers, and there will still be secondary business for the local economy, from having production there (energy, raw materials, etc.).
As the article mentions, the equipment itself will become less capital-intensive over time, which will make it easier to move. It will become easier to improve standards of living in undeveloped countries (barring political restrictions).
The real cost drivers would increasingly become energy, raw materials, and taxes. Large scale manufacturing should migrate to areas with a relative cost advantage for those factors, while lots of manufacturing becomes localized.
The bottom line is that the costs, and cycle times for manufactured goods will drop. Quality of life will improve, even if some improve faster than others.