Review: Ultralearning

I'm of two minds about Ultralearning. The book as a whole is good. It describes various learning techniques that allow you to learn something quickly and use it right away. Most good programmers are ultra-learners: the profession pretty much requires it. In 2010, for instance, I worked through an Adobe InDesign course in 2 weeks so I could layout and write Independent Cycle Touring.

The issue is, I think he oversells the technique. For instance, his claim to fame was that he went through the MIT CS program in a year, effectively "earning" a degree in CS in one quarter the time an MIT graduate would take. But if you looked carefully at his claims, you'll see that he didn't even come close to doing the curriculum. For instance, he skipped the senior thesis. His definition of "passing" a class was to score 50% or higher. (I don't know about you, but that sounds kinda low --- that would be a C average, and you wouldn't cut it as an engineer in Silicon Valley with a C average) He did programming assignments, but it's doubtful if he did any of the serious projects. He probably wouldn't stack up even with an MIT graduate with a C average.

Having said that, I think there's a lot of be said for the tips and techniques in the book. Yes, it's a good idea to have a learning plan. There's also a lot to be said for the intensity of "ultra-learning", which is that you put into practice what you learned right away (which is what I did when writing Independent Cycle Touring). The references to and analysis of spaced learning and how to attack the highest barrier first make for good reading and is worth pursuing. I think I probably wouldn't even have written the above paragraph criticizing the book if he'd titled it "Just-In-Time learning" rather than Ultra-learning, but then he might not have gotten a book deal out of it.

I hesitate to put the "recommended" tag on this book, but it's readable, actually has good tips, and is short enough that you're likely to extract real value from it.