This guy explains that in Harvard, in courses every student has an incredible high level of knowledge, in terms of having read a large amount of papers dedicated to the specific topic.
Now the AI relevant part comes, when he continues that the conversation in class is not about the students personal opinion but about evaluating the quality of the papers they all have in common as base. The form of doing so is that they point our errors, and I assume also qualities, of the papers in discussion to pin down those they will state from and continue working with.
I assume the personal opinion is for the Master or PhD courses.
The AI point is that achieving mastery in science needs training of correct progressive conclusions aka the personal opinion from an early stage on. Scientific progress is based on combining given facts in a creative way, like an Apple and that all is attracted by the Earth' center to discover Gravity as an invisible force.
If his psychological department RAGs an AI on all Harvard papers their teacher can save massive amounts of time using that AI as an assistant for valuing the students evaluation and decision making process to than focus on something they don't do yet, it appears: Telling their opinion in a stage in which they must fail yet can early show their potential for Summa Cum Laude over Quod Erat Demonstrandum.
#cyberpunkcoltoure
