This is the virtuous upward cycle, slow as it may be.
As a formerly underpaid third-worlder (I am now a Canadian) I would’ve LOVED more Western companies coming in to exploit me for my labour.
The only thing "the west" (or rather, the world bank) has done to address poverty is redefine it out of existence.
China has seen an uplift in wealth because they could leverage co-dependence to enforce better conditions. That isn't going to happen everywhere.
When people have more economic opportunities that are rooted in a solid reality, they acquire more of the only real power on this planet - economic power.
Reading the history of the working and middle classes in the West, it becomes clear that as they very gradually got more options over the centuries, they started building various edifices to cement that power. Unions, parties, laws, political traditions.
Next I would argue that human greed is not unique to us but is simply a form of price optimization. Corporations and consumers in the West, or planet Zog for that matter, will always want their sneakers manufactured for ten cents a pair if they can get it.
For that reason, sending these economic options to the third world is a very real and sustainable economic situation. Contrast that with artificial political solutions like regular aid packages (I’m not talking about disaster relief or aid for countries at war like Ukraine, I’m talking about funding some African country because they are poor).
Eventually, a situation similar to middle class life here in the West will develop there. Keep in mind this took centuries here even when the West owned the world.
Eventually, as the sweatshops move from country to country, they will run out of poor people to make sneakers for ten cents a pair. We’ll live in a world where everyone is too rich to accept that job. I want to live in that world asap, and the fastest way to get there is through free trade.
One can try to build an alternative company with all those extra benefits you laid out but it still has to deal with the economic reality of clients like OpenAI not willing to pay more.
If gp's hypothetical Nigeria outsourcing company costs +3x or more to label data than Sama Kenya (https://www.sama.com/) -- because they have to fund those extra benefits, then game theory economics predicts that OpenAI would choose the less expensive Sama Kenya.
This means the hypothetical Nigeria company doesn't have revenue to fund those extra wages and benefits.
Just because a company has an alternative (and more expensive plan) to compensate employees, it doesn't mean customers will pay more for it.