We haven’t found a way yet to prevent authoritarian regimes from arising and spreading, so it’s unclear how AI will save the world. On the contrary, AI will make it easier for authoritarian regimes to expand and maintain their control.

All regimes asymptotically tend towards authoritarian in the long run; from their POV it's just easier to do their job that way. AI will greatly accelerate this trend.

This feels like the correct take to me.

AI will also make it easier for freedom fighters and democracy. If everyone on earth held a four-year bachelors degree, don't you think that would pose an issue for demagogues and emotion-based politicians?

Hopefully it's obvious to all of us that no technology, no matter how neat, will solve the social problems of humanity for us. And this is just the start of how I disagree with the author. But I would say it's not exactly a detriment either

Would it? If everyone on earth did have a 4-year bachelor’s degree, but you lived in a social-credit driven, machine vision, surveillance world, and you’re punished every time you step out of line—what are all those educated people going to do? They can’t even meet or communicate in private

The idea is that educated people would not be as likely to support authoritarian regimes. If you take authoritarian power as a given… idk what’s the point of anything?

I do agree that increased surveillance is very scary. But we have to maintain hope that the future holds promise for the less powerful to influence their government - in the 1770s it was via muskets, and in the 2070s who knows what it will be. (As an aside, I personally doubt it’s going to be guns…)

I don’t know about that. Ideologues who provide the justifications for authoritarian regimes tend to be very educated people. In addition, opposing authoritarian tendencies might require more the qualities of activists and physical action than intellectual education.

How about a world in which bullies who spit on the social contracts (and society in general) get a free ride, benefit from it and win? Would you prefer that?

Weren’t they responding specifically to a comment that said AI would make it easier for “freedom fighters and democracy”? I don’t understand how your questions relate to that. Furthermore, you seem to be implying that AI will help deal with the bully problem… how?

You’ve also given an example that the world in which educated people can’t communicate in private is a poor place.

I’m giving a counter-example that a world in which there are no records and no supervision of antisocial behaviors could also be a poor place.

The way that AI can deal with bullying is exactly the same a good teacher could. You build a model of a good and intelligent teacher that deals well with bullying and put it out. Bullying gets solved. Everyone who is not a bully is slightly better off and can enjoy it.