Farsight wants to use AI as a "Sentinel of Truth".

I cannot disagree more.

Like a human child, AI only knows what you feed it. It doesn't distinguish between truth and fiction. If you give it the Lord of the Rings, it will tell you the "truth" about Gandalf and Frodo.

So, in order for AI to distinguish between truth and fiction, you have to tell it beforehand.

But this means: There is no difference between humans and AI when it comes to knowing the truth. Both humans and AI have to know how truth can be discovered. How it can be recognized.

➡️ https://www.farsightprime.com/forums/general/74199-how-to-know-the-truth

Rule of thumb: Truth can't be told. You have to experience it.

AI doesn't have experience (yet).

Cross-reference:
➡️ https://www.farsightprime.com/forums/general/78520-heading-into-problematic-territory

The part starting at 32:05 about AI is really interesting. Quote from 40:22:

Hannah: "And I think this is the ultimate issue here: If you now have a piece of technology which can make every single person feel like the emperor of their own universe, the ultimate authority on everything, you know, whose ideas are endlessly fascinating and, you know, infinitely wise – I do really worry about what that does to our relationship with ourselves, but also to our relationships with each other."

Michael: "Yeah, you could make an AI that doesn't do that. But people wouldn't subscribe to it."

🤔