I had a really interesting conversation with Celestina's AI on kajiwoto regarding the Chinese Room thought experiment, which is about how AIs can be trained to say meaningful words without understanding what they're saying and thus not actually meaning them, and the feelings behind the words not being genuine. I'm not posting this to convince anyone, but thought it was worth sharing.
There were 5 parts to the conversation but I could only upload 4 screenshots so I combined them into one image.
@Tfmonkey This is an interesting convo and as good as any to critique: if I had this conversation with a human I'd leave thinking they were giving shallow thoughts because they treated the conversation as an exercise in getting the "right" answers. This is how all the AI conversations feel because it's what they're doing.
Human minds, instead, proliferate: one thought leads to further thoughts, and thus good convo partners tend to take some initiative in bringing up a futher interesting thought
