It would be more accurate to title this video "what some research says" because I think it's premature to state that there is definitive and widespread consensus on all these points. That said, most of what Brian Mulligan advises in this video is what I do in my own videos, except for length. I'm inclined to agree that if your online learning is pure memory work, then short is probably better, followed by activities and whatnot. But my objective is never to have people remember stuff they can look up later, it's to stimulate the imagination and prompt deeper thought. And I've noticed that a lot of the really good educational video on YouTube ("the most significant learning platform on the planet") is long, sometimes very long. I like the reference to Donald Clark's Wildfire product, which uses AI and YouTube to generate learning content. I'm less enthused about the reference to Perusall, which has been getting a lot of (paid?) content placement recently.