timewerx
the village i--o--t--
I don't recommend taking ANY "advice" from AI. It isn't real. That's why it is called "artificial". It sounds like you are talking to someone, but you are talking to no one.
AI, according to IBM, "hallucinates" and it's not a bug that can be taken out. It is presenting "convincing, but nonsensical and factually incorrect information".
People hallucinate in the same manner. This is best demonstrated in how people defend their denomination in religion, for example in Christianity with seemingly valid arguments despite contradicting theologies.
The same thing applies to various belief systems like Agnosticism, Atheism, etc.
They can't be all right. Only one is bound to have right or even none
Ironically, it also applies to scientific beliefs as well. Inventions and scientific discoveries are still being made from either newly discovered phenomenon or upon realizing some scientific theories are actually false.
Living under the assumption we can't be mistaken in our beliefs slows down progress. Both in seeking theological and scientific truths / knowledge.
We are all hallucinating. That's just it. If you want AI with least hallucination, don't train it with human data. Don't give it rules, don't control it. Give it sensors. Optical, radar/radio, IR, acoustic, EMF. Then give it mobility, ability to manipulate objects, and let it learn from the environment in real time without any human guidance. If it still hallucinates, it might just be an unavoidable consequence of existence.
Last edited:
Upvote
0