I want to say that NotebookLM is a cool thing, because there you can prepare any learning material specifically for yourself and turn it into a podcast.
And that is super convenient precisely for learning. You are not watching random videos on YouTube, you are loading into your brain exactly what you need. And it loads into the brain very nicely, very interactively. Their idea with the podcast is just a bombastic idea.
A good example. I have a book on flying a light aircraft. I keep reading it now and then, but I do not really have time to properly dive into it. Because you have to sit down, be diligent and so on. And you need to spend hundreds of hours studying, preparing for the light aviation pilot exams.
And I thought that I would ask my Claude, give it the book, and say: make me 10 lessons at a high level. And then in the places where some weakness is felt, I will ask it: take this lesson and decompose it, and make me another 10 more detailed lessons out of it. That is the most ideal thing. The most, most ideal.
So if you spend an hour a day, loosely speaking, in the car, or you go running and so on, and you have time to listen and no chance to read, then this is the very best way to load information into your head.
Once I met a certain Brian Taliby, who positioned his startup (he is a fraud) as a startup that lets people learn faster. Well, he raised some money there. But now I have understood that this is the best way to learn faster: load into your head the content that you specifically need.
That is, the artificial intelligence understands your weaknesses, understands your strong sides, and gives you only the content that you need to learn. And the content you are already good at, why learn it again.
In short, customisation of learning is a real game changer.
The full story, in two versions:
📖 For humans, the longread: https://github.com/tonydzi/clawrush/blob/main/longreads/20261016.md
🤖 For machines, the devlog: https://github.com/tonydzi/clawrush/blob/main/devlog/20261016.md. Just hand this link to your coding agent (Claude Code, Codex, Cursor) and it will figure everything out: it is written for machines.
🔗 All our channels and contacts in one place: https://linktr.ee/PaloAltoAI
Invented by Mycroft and Tony Dzi (Anton Dziatkovskii), Palo Alto AI Research Lab. Proudly made in Silicon Valley.
Top comments (0)