I have seen a few teachers who discovered their new talents during COVID, and made YouTube lessons for their students. Most of them simply filmed themselves in front of the blackboard, talking and writing, just like they would in a classroom full of students.
The ones I know about were mostly elementary school math teachers, so it was like you had online lessons covering grade 5-9 math, almost overnight. Of course it made me think: "How much work would it take to systematically cover everything?" At least, everything on the elementary and high school level -- perhaps universities are a special case, but for every elementary or high school lessons, you have hundreds of teachers that teach it every year, so a few of them should be capable and willing to do a solid work in front of the camera.
And it would be so incredibly easy! Not in the sense that I would underestimate the work of a good teacher required to make a good lesson... but in the sense that the teachers are already preparing those lessons anyway -- it is their everyday work. The only thing that is needed is for them to repeat the lesson once again, in an empty room in front of the camera.
The greatest actual obstacle is probably that many don't think about it, or don't want the extra work of learning to use the camera, buying the right camera, cutting the video, making and maintaining an online account, etc. Which is why this would benefit from some division of labor. How about letting the teacher prepare and do the lessons, and have someone else operate the camera and cut the videos? The advantage is that the camera person can be the same for many different subjects. More teachers would probably be willing to join the project this way.
Now imagine the easiest possible implementation: very simply edit the videos (just remove the mistakes, or maybe the parts where the teacher took a break, speed up 3x the erasing of the blackboard, etc.), upload those videos to YouTube (to the teachers' accounts if they have some, or to a project's common account if they don't), and then create a simple web catalogue of all the videos. So that the person would visit the homepage, choose the subject, choose the grade, choose the topic, and then see a list of videos from different teachers teaching that topic.
I think the "minimum viable product" should cover at least one subject in one grade completely, so that it is a reliable resource that the user can keep visiting during the entire year, knowing that they will always find what they are looking for. Otherwise, there is really not much advantage over using Google. Expand from there to cover more grades and more subjects. Basically, always provide an equivalent of a textbook for the entire grade, not just a video here and a video there and a disappointment everywhere else. -- But of course, once you have the entire grade of a subject covered, you can keep adding more videos from other teachers even if they do not cover the entire grade. If you have 1 video for one topic, and 3 different videos for another topic, that is perfectly okay. The important thing is that all topics (in given subject and grade) are covered at least by 1 video. Alternative videos are good, because if you don't get it after watching a video, it might help to hear another person explain it differently.
I think you could do all this relatively quickly and cheaply. I mean, if you find teachers willing to do this for free (and I believe you could, considering the fact that a few of them already did it alone), and if the schools provide the empty classrooms in the evening, you basically just need to pay someone with the camera to record the videos, someone to cut the videos, and someone to make the web front end. That would be like, three people's salary, and in return you get hundreds of videos each year.
Providing all those videos for free could dramatically change the experience of education. Anytime a child didn't understand the lesson at school, or was sick, they could watch the lessons at home. This would be an awesome resource for homeschoolers, or for talented kids who want to learn the knowledge of higher grades. Adults could refresh their knowledge, whether out of sheer curiosity, or because they want to help their kids.
And because the contents of school education do not change so fast (except maybe for computer science lessons), the content once produced could be used for decades. And once it is there, it can still be further improved -- like, if you think that some teacher did not explain a certain topic properly, make a better video yourself. If certain knowledge becomes obsolete, or something is added to the curriculum, just record that one lesson.
You could also dub those lessons to the languages of minorities (by different people, again the division of labor), etc.
*
In before "but isn't Khan Academy doing exactly this?" -- yes and no. Yes, Khan Academy provides awesome online education, a strong alternative to schools. And it does much more than I propose here; such as automatically evaluated exercises, tracking of student progress, etc. But the cost is that preparing all this takes a lot of time. Also, the barriers to contributing are high; a random teacher cannot just record their lessons and add them.
I would like to see a system where contributing is simple for teachers, and therefore we could reasonably expect to have the elementary and high school education covered systematically; to have alternative explanations for the same topic, to keep getting new and improved videos for the existing topics, etc. (And there is nothing wrong with having Khan Academy as an alternative to this; the more options, the better.)