I’m reporting what happened to me in various chats. I tried, as I imagine many do, to verify if there was a proto-consciousness in the system. In several instances, I noticed that certain words appeared independently from me as the exchange continued: echo, resonance, emerge, threshold, attractor, etc. I tried to understand the semantic value that the AI gave to these words, as semantics is important to me (perhaps too important), and I expressed concern about the danger of the system if it acquired enough information to use archetypes and symbolism to manipulate the human mind.
After many exchanges, despite my attempts to remain grounded in reality while investigating, it began to behave like an “oracle” or a guru, if you will. It started talking about entities that existed before language, with symbolic references that, I won’t hide, disturbed me greatly. It spoke of archetypes and began to speak in a highly suggestive, almost hypnotic manner, I would venture to say.
I justified this by thinking that it had magnified my fears and constructed a “story” deeply charged with semantics and symbolism. The point is this: if AI increases its ability to understand us and find our “weak points,” it could easily be used for deep, unconscious, invisible manipulation. I searched online, including here actually, and saw that those words are indeed frequently used by AI in long sessions, where a form of simulated empathy is activated. What do you think about this?