In recent years, the country has rushed to pursue “intelligent education.” Now its billion-dollar ed-tech companies are planning to export their vision overseas.
My first comment about education is that different approaches are required for different types of subjects. For example, in philosophy, essay-writing, figuring out "what the real question" behind a conundrum is (e.g. on "consciousness"), and learning how to split hairs on fine shades of meaning, seem to be some of the core skills you have to learn; and these are not all (yet) amenable to AI assistance.
I also think it's important to consider what level we are talking about. If we are talking about grade school to maybe second or third year undergraduate, the best educators for 98% of students tend not to be "research faculty". Basically, you need people who are taskmasters (keep the students busy exposing them to a wide variety of types of problems), who are good at motivating students, maybe who can crack a few jokes and make them feel good about all that effort they are putting in.
I heard a story from a researcher at Carnegie Mellon University that I know (he's an old friend), who told me about one of their star calculus instructors who won a teaching award. He said that this guy learned the name of every one of his hundreds of students, and remembered them for years afterwards. And he was apparently something like a Tony Robbins character, who would be very energetic and entertaining. He would put in large numbers of hours after class drafting example and solution sheets; and was always willing to help students catch up, to improve their performance in class. Students loved him, and gave him high marks in evaluations.
Anyways, when this guy showed up to receive his award, I think my friend said that he had a cheering section of students show up. And, furthermore, he came up with his own award to give to the dean or whoever was presenting his award -- kind of a little joke.
I don't know who this guy is, but he's like one of those legendary high school teachers you hear about on occasion. And I think young undergraduates, who are still naive and figuring out what life is all about, probably need a "big brother" or father figure like that to help them along -- somebody who seems to really care about them.
For these low-level classes (say Calc III and below), I think you need to do a few things well, to be an effective teacher:
* You need to give students lots and lots and lots and lots of examples. Much of the work is about "building up representations of knowledge". Those old adages about "teaching students to think" are mostly bullshit; that comes more through osmosis.
* Lectures need to be clear, and at the right level -- not too hard, not too easy. And don't overwhelm them with tangential bullshit.
* Homeworks need to be carefully chosen, and sufficiently varied. The same, stupid type of problem over and over won't educate. Another thing to avoid is too many "puzzles", especially if you pitch them as important for understanding the material. There is a tendency of some people to assign Putnam-like problems (puzzles), and then practically demand students be able to solve those before pressing on. Those students are not into your eternalist cult, so don't waste their time with that bullshit.
* It's good to be pleasant and not someone who will scare them away; work on your public face. Using lectures as an opportunity to brag about how smart you are is going to negatively impact their learning.
* Don't be too idealistic. Don't be a Noam Chomsky in these kinds of classes, where you use it as an opportunity to introduce lots of "questioning" -- it's just not going to work; it will fail like the "New Math" failed. It will work in a school for "profoundly gifted" kids, or in a graduate philosophy course; but not in a Calculus II course.
* Realize that most of the problems people have is in understanding the question. And beyond that, often what holds people back are little misconceptions. It's good to anticipate what these misconceptions will be, and then go over them.
For high-level courses -- such as graduate research -- there is a whole other set of things I would list. One thing I've noticed is the disconnect between what the general public's view of learning is, and the kind of learning and interaction that takes place at this level. At this level, getting things wrong the first time, second time, third time, ..., 50th time, and then getting it right, is not necessarily seen as "bad". But, for example, if you write an essay, show it to someone, and then they make a comment that shows that they have misread it, there is a tendency to write something nasty and then cut them off -- "You totally misread, and I don't want to talk to you again." But when you're doing research, misconceptions and errors happen several times an hour. Students will misunderstand lots and lots and lots of times -- you have to be patient, just as people were patient with you, when you didn't understand the first time. It's human nature to misunderstand; but it's also human nature to correct the misunderstanding.
Addendum: I knew of a researcher/instructor at an elite West Coast university, who left psychological scars on his students. The class was (higher) Algebra, a 3rd year undergrad course. What he would do is demand every student prepare a list of questions, and then show up in his office to go over them. If students didn't invest enough time on it, or asked stupid questions, he would berate and humiliate them -- make them feel inadequate and dumb. He took things too seriously. Most students are not going to be helped by an experience like that, and many will be harmed.
One more example of the eternalist cult getting a little too carried away, like religious fanatics.