The objective function for serving content surely isn't literally to get people angry, rather to "promote engagement" or whatever, which turns out to be the same thing.
Can you ban a company from trying to optimise its key metrics? No. So what do you do? If the enraging content is there anyway, what's wrong with pointing people to it? If you're allowed to sell dynamite to the public in your shop, why not point angry customers to it?
To me the whole model is broken, and picking individual bits of it that aren't is not the constructive approach.
The objective function for serving content surely isn't literally to get people angry, rather to "promote engagement" or whatever, which turns out to be the same thing.
Can you ban a company from trying to optimise its key metrics? No. So what do you do? If the enraging content is there anyway, what's wrong with pointing people to it? If you're allowed to sell dynamite to the public in your shop, why not point angry customers to it?
To me the whole model is broken, and picking individual bits of it that aren't is not the constructive approach.