[AJ007]

The US has basically done this with Cuba and North Korea. Unfortunately whether everyone is better off or not presents a non-falsifiable hypothesis.

I think there certainly should be a distinction between what is being traded. Medicine & building materials is one thing, weapons & censorship tools is another.

Google's problem here is that they are going to have a much more difficult time explaining why x,y, and z are not censored on the search engines of countries that supposedly are democratic. If they are removing nuanced items for China's regime they are going to find themselves being required, legally, to remove all kinds of things in just about everywhere. That will be a net negative to the free world.

My suspicion is that Google expects Chinese compliant censorship is much more feasible now, without an army of workers, due to machine learning. That is very concerning and if true, I think it presents a case that those involved should resign or be removed immediately.

[bobthepanda]

Honestly, they'll probably screw up something Youtube Kids style and then get blocked again, since that's what happened with their human censors several times the last time they tried running a Chinese search engine.

As far as I have seen and that I am aware, AI so far has been really bad at classifying new emergent uses of existing media (e.g., reappropriation of things for memes of a political nature or otherwise.) Are they going to be able to detect the next Winnie the Pooh-type meme lampooning Xi Jinping?