AI makes large scale retroactive thought policing practical. This is terrifying.

I think people will come to the conclusion that they don’t want to post anything online. Private chats, if they exist, will stay but already, if you generate a work of art, you donate it more or less to the shareholders of Google or Meta. This is even before the thought policing and similar implications.

I would argue that's been happening for a long time with much simpler methods. Take this website for example, I get points for posting comments that people upvote. If I post a comment that people don't like, I will get downvoted. I obviously want more points because... so I naturally work to craft the comment that I believe will net the most upvotes. This isn't 100% terrible because this helps weed out the assholes, but I think it has much more insidious effects.

Feel free to call me an accelerationist but I hope AI makes social media so awful that no one wants to use it anymore. My hope is that AI is the cleansing fire that burns down social media so that we can rebuild on fertile soil.

Like search histories but far more.

Roko's basilisk is on it's way.