I think one of the biggest mistakes we have made as an industry is conflating the words “AI” and “LLMs.” The irony is right there on the surface. Naming is one of the hardest things to do in software, and we’ve done it poorly for the primary tool of software.
Sibling comment is great for PBR. For KS, unfortunately most schools have a blind spot around it. This is partially because KS directly implies the legendary Bell’s theorem, which is experimentally testable and has allowed folks to prolong their grief over objective reality by bargaining with those experiments rather than doing maths; teaching Bell’s theorem is less of a headache while still breaking classical assumptions. It’s also partially because KS is not well-understood as a matter of folklore. I would say that maybe nLab’s page on KS states the important part, but they still hedge by using lots of formal language. Here’s an informal consequence of KS:
Given. We exist in three spatial dimensions.
Given. Photons of the Standard Model have spin 1.
Corollary (QM is indefinite). The spin of a photon isn’t definite prior to measurement.
This leads to the infamous Kochen-Conway theorem, better known as the “free will” theorem: if humans have free will then photons have free will. Kochen has been insistent that this family of results is solid and wants his peers to pay attention; he releases a position paper every few years on the topic, e.g. Kochen 2017 or Kochen 2022.