5 ms·
Everybody has a different idea of AI safety. There is the Skynet-scenario (don't let AI take over, by hacking or influencing humans). And there is the thought p
by tjansen 3y ago
Everybody has a different idea of AI safety. There is the Skynet-scenario (don't let AI take over, by hacking or influencing humans). And there is the thought police (ask ChatGPT to say something nice about Donald Trump vs something nice about Joe Biden; https://philip.greenspun.com/blog/2023/03/05/chatgpt-waxes-poetic-about-joe-biden-and-corn-pop/ https://philip.greenspun.com/blog/2023/03/05/chatgpt-waxes-p... ).