@UlrikeHahn There is a lot of value in researching the psychology of the "AI"-Human feedback loop; not by using LLMs as a model of our mind but by analyzing what we believe about LLMs (or "agentic systems" based on them) when exposed to it; a skinner-box we collectively create while using it.

We may have a chance to expose meta-mechanisms of the social, laws that govern the probability of socio-psychological effects in a way that's open to epistemological practice.