AI 摘要
大多数对齐计划: 第一步)创造沙神 第二步)... 😈 欺骗沙神 😈 ... 第三步)沙神永远保持忠诚仆从 "当前的对齐工作都只是给修格斯涂口红。" -@romanyam
Most alignment plans:
Step 1) Create sand gods Step 2) … 😈 Trick the sand gods 😈 … Step 3) Sand gods remain loyal servants, forever
"Current alignment work is all about putting lipstick on a shoggoth." -@romanyam
I don't know who needs to hear this but preventing the models from learning about the tree of the knowledge of good and evil is not a good alignment strategy.