What everyone complaining about what OpenAI can "afford" is missing is that if OpenAI were to pay more, the only rational thing for them to do would be to move the jobs to another country that has lower risk levels. This would leave Kenya without the jobs at all.

Thinking you can have both higher wages and jobs in unstable countries is either advocating for state-mandated outsourcing or is wishful thinking.

Ethics and current day globalized economics don't always mix well. It would still be the right thing to do.

If a company acts unethically that is bad, even if it's in the company's rational self-interest. It rings especially hollow when the company's supposed mission is to "ensure AI benefits all of humanity"

> Less bad exploitation is still exploitation.

Your statement is problematic, because (a) clearly at some point "less bad" stops being what anyone defines as exploitation, and (b) if you're not living in the tradeoff space, you're not qualified to judge where that crossover is.

There certainly is exploitation to be found*, but paying a multiple of all other opportunities, even if that multiple is below what is available in a different context, is not exploitation.

It's putting a thumb on the scale, accelerating the tilt towards equity.

* There are articles about labeling or screening done by workers in USA that are as PTSD-inducing as the linked stories, where people were paid US wage and had to quit. Don't confuse the job with the pay, these are two different issues.

// Disclosure: I grew up in central Africa. In my teens I worked in the local tech economy, not the expat economy.

Irregardless isn’t a word