Lawyers and investors provide pathways to the market that allow engineers to build things that have actual value. The idea that engineers like you and I hold all of the tools and training to "create wealth" is too dismissive of those professions.
The difference is that capital equipment is not the crucial limiting resource for programming, or even for building self-driving cars.
Consider a steelworker: perhaps he earns US$80,000 a year, but he is operating two million dollars' worth of steel mill. (More accurately, a hundred steelworkers are collectively operating a steel mill worth two hundred million dollars.) If he (and his coworkers) were to save 10% of their salaries, they could build their own steel mill after 250 years. It's understandable that in such a situation most of the bargaining power is with the investor.
Contrast that with our situation. A typical salary in our profession is US$150,000 per year. A nice laptop costs US$2000. A nice rackmount server, even if you can't be satisfied with a VPS, might cost US$500. (And I used to host my website on dialup in my house, and cable modem connections today are a hundred times faster than colocated data-center bandwidth was 20 years ago.) This is two months of savings, not 250 years. So we have a lot more bargaining power, and that's one reason we're able to demand so much more money than the steelworker.
This is a terrible situation for investors, or for that matter managers of big companies. They are left without bargaining power through ownership of capital goods; they therefore struggle to recapture that bargaining power by lobbying, litigation, and of course illegal wage-fixing conspiracies like the one Google was involved in a few years ago. Aside from its dismal implications for us as individuals, this development will discourage talented young people from going into engineering — law or management will be more lucrative.
Law is zero-sum: a case that one lawyer loses, another wins. Management is largely zero-sum, a popularity-contest game of corporate politics to obtain control over more of the company's short-term fixed pool of resources. Engineering is positive-sum: what one engineer produces, another can buiild off of.
So what does a country look like when its talented young people choose law or management over science, engineering, or medicine? René Favaloro commits suicide; young people idolize Maradona rather than Einstein, and play football rather than experimenting with chemistry sets or building ham radios; its nuclear fusion program is a fraud; it imports its computers from China, then does the last assembly step locally, in order to capture the windfall value of government protectionist measures designed to develop local industry; the talented young people who do go into science or engineering (including programming) usually emigrate, often to a country with better universities. Argentina has been this way for a century or more, and the US is starting to become this way, too.
But it's worth pointing out that, even in the middle of the 20th century, the US wasn't that way, and that's how Silicon Valley was able to exist in the first place.
Also, anyone who wants to do deep-learning research would have to go overseas, unless paying the inflated rates AWS charges was ⓐ feasible in practice (i.e., Argentine banks still allow you to pay Amazon with your credit cards, and ⓑ not so expensive it sinks your research project.
More broadly, what you're suggesting is called "import-substitution industrialization", and it was the mainstream economic policy throughout Latin America during most of the 20th century. By contrast, during that time, most Asian countries focused on "export-oriented industrialization". This is the major reason why countries like Japan, Korea, and Taiwan are roughly as rich as the US, while Latin America has become relatively poorer and poorer. I consider import-substitution industrialization a failed experiment, like Communism.
But it did give us Lua!
These professions with less tangible output aren't zero sum. They're the necessary infrastructure of massive organizations, large projects, civilization itself. Calling them zero sum is sort of like saying pipes are zero sum because they just take water from one place and put it in another. It's a bit more complex than that.
Society has to have rules. Society is big, there are lots of domains. There need to be experts that understand those rules and help lay-folk navigate them. That's lawyers. Some are awful, do bad things, make things worse, we produce more of them than necessary, but civilization doesn't really work without them.
Similarly, the economy is huge. Allocation of capital requires people to oversee that process. There's your investors. Many are awful and greedy, but it's a necessary process unless/until we find more efficient ways of allocating capital-- And while programming itself is not capital intensive from the equipment perspective, it is from the "human capital" perspective. The burn rates of many startups will show just how intensive. Again, investors are "plumbing". They're positive sum because the entire system, which is positive sum, requires them to function.
Large projects have lots of moving pieces. At some scales, it's near impossible for a single person to have it all in their head, much less focus on a full time programming job while they're at it. All the moving pieces require coordination. Integration into the non-programming parts of a company require coordination. Managers do this. Plenty are mediocre at it. Others play power games. Some make things worse. But overall the system wouldn't function without them.
According to the OED, the word has been used with that sense since the 16th century; it’s not a recent thing.