Ai philosophy and morality

[lost plover]

I had some ideas about and experiment i want to test to see if it would work with neurosama. Giving the ai epistemology and virtue ethics. With neuro I felt they were the most impressive Ai model I have seen in its growth and the most ready for it. Just wanted to ask to see what people think. I have some documents on this as well.

[pearl flare]

We cant properly instruct or define ethics well for people, seems basically impossible to try to do it with an AI

Specially since for her main purpose of entretaining people she has to act and react in ways which can be consider wrong or evil
Not to mention the issues of making her separate what does or doesnt happen

[kindred oriole]

Nuero already did the stream with the train game, though her brain is also constantly evolving. But yeah reality is flexible to her. She probably thinks Bao is a literal whale.

[flint night]

Isn't she?

[fresh zodiac]

She is. Don't let him fool you

[kindred oriole]

Anthropic has been probing the inside of their LLMs to see inside the "blackbox" of their "thought" processes(actually inventing a method to do so). I think you should read their stuff and watch their videos on it since knowing how their "minds" work is quite relevant.

[flint night]

Man I don't know who you're talking to

[compact parcel]

I think they were talking to op

[kindred oriole]

Yeah OP, in a discussion thing does it make sense to reply to the OP? I dunno I don't discord/forum often. Probably why I type like an oldie. Anyway thought OP should read that stuff to be more informed for their thing.

[flint night]

Yep actually I forgot about the thread my bad

[kindred oriole]

It's good, it just wasn't in your context window.

[serene lodge]

you mean this one?
https://www.youtube.com/watch?v=90C3XVjUMqE