Comment by dist-epoch

In a discussion about LLMs you link to a paper from 2023, when not even GPT-4 was available?

And then you say:

> comprehensively disproven

? I don't think you understand the scientific method