> So, what would it take for that future not to happen?

First, the obvious one: war.

Conflict involving Russia, the US and China in some combination, is by far the greatest risk and will remain so indefinitely. It would throw everything up in the air.

There is some serious risk from eg Pakistan, a nuclear-armed country that will surpass the US in population in the coming decades and currently has a GDP per capita of $1,500. That level of poverty in a country with so many people, with nuclear weapons, and rather lacking in political stability, is just a huge risk over time. It borders billions of people, including two nuclear powers in China + India (both critical pieces to the present and future global economy). There's no indication that Pakistan is going to develop economically any time soon, so the nuclear nutso risk there will remain immense, primarily from black swan events within the nation. It's beneficial that the US is geographically far away from China and Russia. In the case of India + Pakistan + China being so close, it's concerning.

North Korea is a modest, although serious, risk. Heavily regional in nature, with some limited threat to the US mainland. That goes back to the US - China war risk problem mostly, where crazy events can spiral. China is going to want Taiwan back formally within the next two decades, I think that's almost guaranteed. Their patience is already exhausted, when it comes to what they view as US tampering with Taiwan and their local interests (every day that China becomes stronger in the region in terms of their power projection, they will tolerate the US-Taiwan situation less).

Saudi Arabia and Iran will push toward the development of nuclear weapons. Nothing good is likely to come from that, and it would plausibly risk drawing in all the great powers one way or another. The US side may do something crazy to Iran (try to Syria'ize it) to try to stop its inevitable progress toward nukes.

There are some serious population + poverty risks in Africa. Their economic development needs to accelerate dramatically over the coming decades, to keep up with the huge population growth (immigration flows could severely rock Europe otherwise). Large scale wars are probably a risk without enough economic development there. That's quite localized to Africa, rather than being a US-China global war matter, however it's still serious. Africa will be a large share of the global economy almost no matter what, due to their rapidly growing population figures. It's not much of a risk issue for the next 20 years (it particularly is ~50 years out however).

On the economic front, you could see a major central bank or two make a historic mistake in some direction (primarily the US, China, EU), sparking a great depression risk. It's unlikely near-term, their current approach can be maintained for at least the next few decades, assuming some placating toward the lower economic classes to keep them from outright revolt (the economic transfers have to go up in most cases). I'd expect the rich to placate when they feel they have to, they have a lot of economic room they can afford to give yet.

The huge debt levels - and the allocation of capital en masse to that - in the larger economies is keeping a heavy lid on most inflationary forces (you're seeing asset inflation however, which is legitimately dangerous just as major consumer goods inflation is). It's the Japan scenario, playing out in the US, EU and increasingly China. Debt is soaking up tremendous amounts of capital - and spitting back very little yield - that would otherwise be sloshing around these economies sparking much worse inflationary forces. I don't know if the Fed is stupid or playing stupid, it's obvious why traditional inflation is subdued despite a decade of low rates (again, it's for the exact same reason Japan has seen such little traditional inflation for two decades: debt is occupying increasingly large swaths of productive capital, at ever lower rates of return). That can't last forever, what you end up with is an alternative inflation outcome, namely sharper currency debasement as debt creeps too high to sustain the gradual QE approach (which results in a similar ultimate hammer to standards of living, usually more sudden).

In the next two decades most developed nations will add net wealth equal to or greater than the debt they need to debase to stay above water. The US for example will soon need to do the equivalent of QE'ing a trillion dollars per year (due to the perpetual large budget deficits that can now never be turned back), it can easily add more net national wealth than that per year. It can either do that QE gradually or in bursts when the shit hits the fan from time to time (recessions are ideal for them to do it, as it provides a good cover excuse; so I'd expect high aggression on that front from the Fed in each recessionary period, they'll use it to hammer down govt debt costs with each cycle). The primary issue, is that the bulk of that net added wealth will continue to allocate toward the top, causing ever increasing friction in society (requiring greater income & wealth transfers, and further development of the US social safety net, including large moves on healthcare and education).

I view the trade conflicts as a very modest risk, short of outright Smoot–Hawley behavior in the major economies. Could happen, unlikely though. The world can deal with persistent trade conflict between China and the US (with various nations caught in the middle in different ways at different times). It's the next level up that is to be watched for, where global trade risks an actual crash.

This era of quasi anti-globalism and increased nationalism will produce sudden jumps in technology, as tensions rise and competition rises between great powers. I'd expect it to last for decades more yet. The problem is that the risk of large military conflicts goes up with that. Entrepreneurs will produce spectacular new leaps forward in many fields, from AI to aerospace to biotech, after a bit of a boring couple decades in most fields. Major governments will pour more resources into R&D, as during the cold war era, prompting some sudden leaps forward that might have otherwise taken longer. China will probably single-handedly cause the US and EU to get back into fighting shape in that regard, after a few rather fat & lazy decades.

> Or what should the people that are going to be the most affected do?

Develop skills that leverage; skills that multiply radically rather than add modestly. Or otherwise develop trade skills that benefit from the top economic classes getting richer on what's going to happen. Output per capita overall will continue to rise, AI & robotics will push that to the next level. Humanity can no longer develop output to a far greater degree without shifting to the next era of automation, where we hand it off. We've heavily maxed out our potential, which is why we're struggling so mightily to move the needle on productivity in developed economies (despite having all the capital & technical capability theoretically necessary). We're pushing on the ceiling. AI et al. will take over from here (in the coming century), enabling future outcomes - levels of productivity and output - that we can't otherwise accomplish. Robotics that output 2x or 5x more per shift, 24/7. It'll largely keep a lid on inflation in most consumer products in the developed world. Most of the new output gains will be shifted to fulfilling the demand that will rise out of Africa, Latin America and parts of Asia (nations / regions that will see huge population increases and corresponding consumer demand). Three billion more people will want shoes, food, electronics, Internet services, medicine, healthcare services, et al.

Highly educated, stable, smaller population nations will be the biggest winners. They'll more easily outrun their population's welfare demands (a serious per capita problem, with slow population expansion). They will more easily be able to focus their limited government resources on maximizing positive outcomes, and those positive outcomes will have far higher multipliers on them in the future in terms of output, wealth creation. We should be able to do some interesting output leaps, where small nations could punch very far above their weight on certain manufacturing. Very large amounts of human labor will be a net negative in most cases. Canada, Australia, New Zealand, the Baltics, Scandinavia, Ireland, Netherlands, Austria, Israel, South Korea, Singapore, will do particularly well (a lot of South Korea's easy upside is already in the tank though, their progress will continue just more slowly). The Baltics will become Scandinavia-lite, visit them now before they become affluent and everything gets a lot more expensive there. Czech, Slovakia, Slovenia, and one or two others in that region will probably do well. I think parts of Eastern Europe could see good gains, where they push into the upper tier of the middle income bracket; Romania for example might do that. Vietnam and a few other Asian nations that take some of China's manufacturing as a risk diversification, will do relatively well as capital continues to flood in (assuming they don't regress politically, Vietnam is a risk for that). In Latin America, I think Colombia could be very interesting for the next 20 years as a potential to break-out economically.

Reshoring some manufacturing - in the US, EU and elsewhere - will continue to happen, for expediency (and possibly strategic reasons, such as IP protection or national security). It'll vary heavily by the manufacturing in question. China will lose some of its overall manufacturing crown (to reshoring and general competition), as the US previously did. That is, they've already peaked on that front. However China will also continue to push up the ladder on more advanced production, whether in biotech or aerospace (Boeing and Airbus will lose most of the Chinese market).

It'll be great to be a plumber in the next 20 years, no joke. Or to run a plumbing service business. Certain manual labor skill trades will do very, very well. The top 50% in most developed nations will continue to grow wealthier overall, and plumbers (and other skilled trades) will command more accordingly. It'll be terrible to be a farmer however, the robots will eat 3/4 of what's left of that labor market.