The way to leverage best LLMs at schools and research
Since one year already, or just a bit more, the educational sector is flooded with AI generated home-works, in which there is a degree of acceptance among instructors. It is assumed that at least the work was read, if not conceptually conceived.
There are domains that can leverage the better approach: instead of copying generated content and read it, there is the possibility to treat the LLM as a interlocutor agent, in which one can actually engage in the thematic dialogue.
It is the case in philosophy for example, in which with the help of the LLM agent, one can create helpful dialogues, that certainly represent a far more richer and desired experience than one-shoots home-work only reading (and correcting maybe) approach.
In this way also our Wittgenstein @ EM website tries to apply this principle, in which there are dialogues, extracted from the prompts with an LLM primarily. A dialogue as in an educational way advocated here: one goes deep and as wide as possible to extract the theme of interest as detailed, coherent or even sometimes as creatively as possible.
To note that creativity here is not the LLMs side: one can't expect that to happen - there is a lot of research in this direction that assesses the limitation of the LLMs as creative, novelty output agents - there is very little “generative” in them. But there is some part interesting from the prompting side: the dialogues potential, on the other hand, and the clever leverage of the “confabulation” aspect of the LLMs, in which, in their moments of “hallucinatory creativity” they may spark something interesting in the advised human interlocutor.
Here an article that actually supports our motif for this website. It sets the working approach that would be more useful in schools and also the way websites like wittgenstein.essentiamundi.com may have a reason to fulfill a need to a further real research side (humanly guided and content 100%) like ai.essentiamundi.com.
_____________________________________
Original paper and excerpt:
https://philpapers.org/rec/SMIRTP-8
“We propose that instructors shift the emphasis of their assignments from philosophy papers to “LLM dialogues”: philosophical conversations between the student and an LLM. We describe our experience with using these types of assignments over the past several semesters. We argue that, far from undermining quality philosophical instruction, LLMs allow us to teach philosophy more effectively than was possible before.”