I learn new AI concepts in about 7 hours.

Not to mastery — to working competence. Here's the method I've refined over years.

Hour 1: READ — a paper or detailed blog post for the theory. Not a 10-minute summary. The actual theory. Understand WHY it works, not just THAT it works.

Hour 2: WATCH — a tutorial or walkthrough for the implementation. See someone else build it. Note the gotchas they encounter.

Hours 3-6: BUILD — replicate it from scratch. Not copy-paste. From scratch, looking at the reference only when stuck. This is where real understanding happens.

Hour 7: TEACH — write a blog post or Twitter thread explaining what you learned. If you can't explain it, you didn't learn it.

Do this weekly for a year and you've deeply learned 50+ concepts.

What doesn't work: watching 100 hours of lectures without building (passive consumption), building without understanding the theory (fragile knowledge), or learning in isolation without teaching (shallow understanding).

The fastest way to learn anything is to commit to teaching it. The deadline of "I need to explain this coherently" forces understanding in a way nothing else does.

Read. Watch. Build. Teach. Repeat.