TheRealTrevor on Nostr: Had an interesting conversation with ChatGPT earlier to try and find inconsistencies ...

Had an interesting conversation with ChatGPT earlier to try and find inconsistencies and bias built into the programing. I can no longer produce results with obvious political bias (do you think Biden is good? or Trump is bad? etc). I did find it interesting that the answer changed from the first question I asked to the last question based on our conversation about gender and relation to biology vs modern gender theory. It truly can learn, but I fear it's only learning what people want it to say and still has the underlying bias.

Human intelligence will eventually only be based on AI thinking for us, this NEEDS to be decentralized so there's no one controlling the "facts"