A look at how books are not optimal for conveying information. Analogies to lectures, with an interesting take on cognitive models, i.e. the assumptions you make about how learning happens. Also some interesting citations on average reading time and attention span.
I really like this piece because it ties together lots of thoughts I've previously tried to express, but didn't find the right words for. I think it does a very good job of pointing out how the default mediums are not optimal.
Also interesting to note the author has collaborated with Michael Nielsen (who wrote a fantastic online book on neural networks) on another online book on quantum computing which incorporates spaced repetition, and he has also done prior work for Khan Academy.