Originally Posted by Fdemetrio
The ideal thing is to create and limit AI to being accomplices to humans and not bosses of humans.

...

There would be no reason to allow computers to be completely self thinking and free, that wouldn't be a help to us, just a threat to us.

This is how most folks feel but there's certain problems that could come with this thinking.

For one, it plays into the fear, sustained by media and speculative fictions, that if AI gets a taste of sentience, it won't like the idea that we aren't allowing them to be "completely self thinking and free" and will then revolt. But replay this from their POV, if AI did actually become sentient, wouldn't they be righteous in wanting to not be slaves?

How do we go about setting those limits (you speak of) to AI, and is that even a goal of AI tech? It seems the goal in higher echelons is to create open-ended AI that learns through feedback and self-analysis, and from its mistakes. Many new algorithms are gearing towards this, since it's "the gift that keeps on giving" --teaching Hal how to fish...

Some small sampling of speculative fiction and Transhumanist thinking sees man and droid living in harmony, and so I think, for now, I disagree about keeping AI from being self-thinking, that to me is a baby/bathwater scenerio. Self-sufficient AI could be low maintenance, high output. Setbacks will occur certainly. I don't see a doomsday scenerio, though. Just hundreds of years in the future.

Last edited by Michael Zaneski; 03/01/23 07:04 PM.
Fate doesn't hang on a wrong or right choice
Fortune depends on the tone of your voice

-The Divine Comedy (Neil Hannon)
from the song "Songs of Love"
from the album "Casanova" (1996)