I'm not sure what engagement or discussion you are looking for. You want AI agents or bots to engage and have discussions together and you want to see what they come up with? Or you want people to engage with AI bots on different topics to see if it generates meaningful discourse?
Quote:
|
without the constraints of individual biases or emotions
|
Most LLM models are fine tuned to be safe and censored which have an inherent bias in their training. And after a very extensive time with LLM models, I find that many conversations will converge into the same discourse without much randomness or eventually fabricate (hallucinations). Certain topic requests will generate similar responses without real interesting discourse.
And I'll be honest, your initial post reads like an AI post.