One of the recent NeurIPS best paper recipients is relevant here: https://openreview.net/forum?id=saDOrrnNTz https://openreview.net/forum?id=saDOrrnNTz > an ex

One of the recent NeurIPS best paper recipients is relevant here: https://openreview.net/forum?id=saDOrrnNTz https://openreview.net/forum?id=saDOrrnNTz > an extensive empirical study across more than 70 models, revealing the Artificial Hivemind effect: pronounced intra- and inter-model homogenization So the inter-model variety will be exeptionally low. Users of LLMs will intuitively know this already, of course.