What a world!

(Putting aside the fact that AI may lead to our demise fairly soon)

Man, what an incredible time to be alive, that I have have a phone call with an AI, whilst on an evening walk around a small English village, with tiny earbuds in my ears so I’m hands free and totally at ease.

I think the 80/20 of my learning stack is definitely anki + AI phone calls (and AI for generating (imperfect but time-saving) flashcards, now that I think about it). Obsidian and note-taking in general is still mostly a distraction, something I don’t actually do that much, kind of busy-work compared to just getting good flashcards made and into my brain.

But, just putting flashcards into your brain isn’t really enough. It gets you a lot of the way there, but AI walk-and-talks play a CRUCIAL role and I’m so excited by them

Co-creation of knowledge

The cognitive scientist Michelene Chi is best known for her ICAP framework, which ranks learning approaches from most to least useful.

Interactive > constructive > active > passive.

  • Passive = just listening to a lecture.
  • Active = taking notes, highlighting.
  • Constructive = creating something, solo. A summary diagram, or engaging in the “self-explanation effect”, where you talk to yourself about what you’re learning.
  • Interactive = co-creation, e.g. via a debate, a dialogue, a group project.

I’ve been doing the self-explanation effect for a while – I’ll go on an evening walk and talk to myself about what I’ve been learning, what I’ve been making flashcards about.

“Ok, I’m gonna talk about what I remember about John Boyd’s zeitgeist. There’s complexity science, chaos theory, there’s Godel’s Incompleteness Theorem and Heisenberg and entropy. So what I know about complexity science is…”

This rules, but you know what’s even better? Having an AI who you can brain dump at, and then who quizzes you.

E.g., just now I was on a walk where I was having AI (and btw, it’s 22:30, I feel like my writing style is abysmal right now, I’m just brain dumping here) I was having AI quiz me about John Vervaeke’s 4 Ps of knowledge, which I only vaguely knew about until today (I’ve spent 3 hours learning about it, making flashcards and reading etc).

The 4 Ps are propositional, procedural, perspectival and participatory. Super profound for various reasons that I won’t get into right now, I imagine I will at some point.

So I was like “I’ve been learning about this today, I’m gonna brain dump at you, then can you quiz me”

And from it quizzing me, I got 2 big benefits:

  1. I only started learning about this today, so the flashcards are brand new, and then knowledge isn’t “mature” in my head yet - it’s still shakey, hard to recall. But the practice of having a conversation about it really speeds up the process of maturing the knowledge, of embedding it. I imagine I’ll have a much easier time with the flashcards when they come up again, because I’ve had a constructive dialogue about this stuff, out-loud, on a walk, embodied, situated, etc.
  2. I got to uncover things that I still don’t understand, and have the AI help me understand them better

On point 2 – I realised that I still didn’t really get the difference between perspectival and participatory knowing. And it helped me with this, and I had a huge eureka moment re: perspectival knowing and how it actually relates to a question I’ve been chewing on (in a very casual way) for a while now. So so cool!!!

So that’s the post. Obviously if you have people in your life who are learning the same stuff as you (e.g. if you’re at uni, or a software developer at an in person job or something) then talking to a human would be even better. but then again, I got to ramble for 30 minutes about all the stuff that I’m currently interested in (that is, I got to fold in some John Boyd and Heidegger and complex systems stuff, and pivotted to asking for career advice near the end) – obviously, an AI has infinite patience for you, which rules.

I took my thinkpad apart today

I spent probably over and hour taking this thinkpad apart today to clean the CPU fan, as I figured it was probably full of dust for 13 years of life. Turns out it was actually pretty damn clean, lol. But it was very satisfying to do so it’s all good.

Turns out I’d been using anki wrong

Turns out I’m in the 10% of people who use “hard” as “I got that wrong”, whereas actually “hard” is meant to be a passing grade. Luckily the FSRS add-on has a feature to fix this, because it was giving me insane intervals (like “either see this card again in 20 minutes or 16 days”, for brand new cards, which was ridiculous). Relevant reddit thread

Where did the day go?

3 hours learning about Vervaeke’s 4 Ps, 1 hour watching a lecture about complex adaptive systems by _ Snowden who it turns out is the Cynefin guy so I definitely want to learn more about him!!! Tasshin has a post about Cynefin, very cool. He has a post about Boyd too, Tasshin knows his shit.

The Smile

Man, I slept on the Smile for far too long. I love Radiohead, turns out The Smile absolutely rule too.