fast.ai courses were the only courses that went a little meta and questioned learning, and how to learn. This developed my interest in the topic Learning How To Learn.
I have spent the past few weeks learning how to learn. Read David Perkins and Paul Lockhart and other stuff under the umbrella.
The fast.ai integrates these philosophies in their course and it is sometimes shown to us and sometimes kept hidden. Just going through the course from start to finish will boost our learning.
If I wanted to learn something new on my own (astrophysics, or painting, or music, etc) and wish to design a method of learning based on fast.ai approach. How can I go about it?
I have gotten some crucial principles from Perkins and Lockhart but am interested to know more around this theme so that I find an empowering technique for life long self-learning of anything.
In lesson 8 Jeremy recommended Meta Learning: How To Learn Deep Learning And Thrive In The Digital World (link). It’s more specific to learning ML, but I think some of the concepts in it really represent the “FastAI way” of doing things and general tips to accelerate your learning.
Jeremy really does have a lot of good tips on how to learn throughout the videos…
And now the first half of Lesson 11 from part 2 which has just been uploaded to Youtube and is keeping me up past my bedtime has amazing advice on how to read academic papers.
Another person I take a lot of inspiration from is Michael Nielsen who has also done a good bit of work to make Deep Learning accessible for all, as well as other topics like Quantum Computing.
Michael Nielsen also seems to have done a lot of research on productivity and learning. He has written several interesting posts under the “Tools for Thought” section on his website.