*) "emulgate", maybe better emulsify, but this is a bit lateral to this? The point being, a homogenous preparation, which is more a superficial operation than an essential one, as the establishing ingredients remain the untouched.

> AI tools are pure rationalists, and are solely capable of reasoning.

Mind that the world isn't in the language, nor our connection with the world. (We know this for about 120 years, since we expelled the referens from linguistics.) Which brings us back to the synthetic judgement a priori… You may emulate this, as a superficial trait drawn from other traces of communication, but it's not what this is all about. E.g., I wouldn't expect true "lateral thinking" from an LLM output.

> My biggest fear from AI…

I'd add to this, it's not just empirical vs rationalist epistemology, it's also about empathy, anything referring to the conditio humana, which is really what any text is about, even a scientific one (why is it that we do want to know, what are the motivations, the regulating circumstances, etc.?).