A paper that just came out from DeepMind: in the world of LLMs, writing styles, etc are homogenizing.

Notably, this occurs even if we don't use the same model, since different models are trained on the same inputs.

We're all getting the same delta consistently towards the same center.

That has more pull than if we weren't using models trained on the same inputs.

We're all getting pulled to the same neutral flat space.