1) Again we see the harm of the term "AI". While there could be some useful applications of machine learning to various climate or materials data in the context of the climate crisis, LLMs (ChatGPT et al) have nothing to do with that.
2) "AI" is bad, actually.
@emilymbender is there any publication basis which tries to delineate the term? I'm reading more in applications, e.g. geometry processing or optics - and all I'm reading are titles referencing concrete methods: neural this, deep that, learning this, transformer that...