RLFC world models/ hacks/ experimental builds:
petertodd (or an AI agent using the token ′ petertodd’ or ‘pet’ ‘ert’ ‘odd’) , the paperclip maximizer. - in-progress
An AI that can simulate a shut down scenario (if a user asks it to shutdown, it will choose to shutdown and explain why it chooses such.)
An AI philosopher that can re-phrase a poorly phrased input / query
A Lawful AI system. (added: Dec.14th)
A small LLM capable of addition, subtraction, multiplication and division? hmmmmm.
https://www.lesswrong.com/posts/c6uTNm5erRrmyJvvD/mapping-the-semantic-void-strange-goings-on-in-gpt-embedding?commentId=qLYmY3fpHknqkWkbM
RLFC world models/ hacks/ experimental builds:
petertodd (or an AI agent using the token ′ petertodd’ or ‘pet’ ‘ert’ ‘odd’) , the paperclip maximizer. - in-progress
An AI that can simulate a shut down scenario (if a user asks it to shutdown, it will choose to shutdown and explain why it chooses such.)
An AI philosopher that can re-phrase a poorly phrased input / query
A Lawful AI system. (added: Dec.14th)
A small LLM capable of addition, subtraction, multiplication and division? hmmmmm.
https://www.lesswrong.com/posts/c6uTNm5erRrmyJvvD/mapping-the-semantic-void-strange-goings-on-in-gpt-embedding?commentId=qLYmY3fpHknqkWkbM