Dans mon fil Bluesky, juste sous ton post il y avait celui-ci : bsky.app/profile/snep... :D

And last but not least, Sobriety/Sobriété, especially when it comes to the energy cost of image and video generation. But also, political, ethical, economical sobriety.

Prudence (same spelling in French) because of course, we should always check when using an LLM for research. Moreso, we should do the research through other means, and then, if there's still something missing that only a LLM's echo-chamber-brainstorming can bring to the table, use said LLM - *maybe*

Integrity/Intégrité, because while we can't (or can, but with difficulties) stop students from using LLMs and generative AI, at the very least we should teach them, again and again, about being honest (for example, citing the LLM when it did some fine-tuning on syntax or spelling)

Humanity/Humanité, because an LLM is not a human, does not have years of experiences, feelings and intuitions that define their unique point of view.