Re the first #3, Jason Yuan and Sam Whitmore just released Dot today, an AI assistant with longterm memory.
What fiction, if any, have you found to be compelling in the recent years of your life? (And do you have a sense of what made it compelling?)
As an aside: I'm curious if you've read The Waves by Virginia Woolf, and if so, what you thought of it.
Robin taught me the tool of "take lots of liveblogging notes".
aw, a little h/t to me.
Just to note, I'd describe the thing I suggested differently:
(mostly I'm going to write this for my own clarity)
your goal at the time of my suggestion was [learning about transformers], and you were primarily just reading and occasionally taking notes seemingly-for-memory (like writing down key concepts). It seemed to me that to "be in touch with the territory" of learning, you should be watching your learning process, ie the felt-sense of the building of the schema as you go over the material, the sensations of uncertainty/ambiguity/confusion, the sensations of insight, etc. To see what you're attending to and how you're attending to it within the constraints of your sequential, moment-to-moment experience. And yeah, a great way to do that is write down what's going on internally while you interact with the material.
(perhaps you didn't hear it this way, but that's a bit of the generator / what I meant~)
Perhaps you are thinking of this (i think) autobiographical essay by Tim Rogers? He also talks about it in his 5th chapter of his boku no natsuyasumi review.
I have an exercise that might help with the phenomenological snapshot (though it's getting at a slightly different thing). Less about examining all sense data available, and more about examining what concepts/shapes are currently salient/present (which may include lower level perceptions as well, if that is the kind of thing that is most salient). I wonder if it might be a little easier.
(Hmu if you're down to try it out? I'd like to run it some more before solidifying it in writing.)
To bring up a specific instance of this kind of problem: that lw post on open/active curiosity absolutely devastated my ability to think about curiosity for no less than a month. Every time I'd prompt myself to think about curiosity, my thoughts would flow toward the "open/active" concept shapes; I didn't know how to stop it (and I very much wanted to stop it. I was frustrated, found the shapes of to be misconfigured, a poor fit. I couldn't access my previous thought configurations on the topic, as they were temporarily overwritten).
The only defense I found in the end was to stop prompting myself on the topic; it took about a month for the shapes to fade, for the forgetting to naturally occur. (I've long thought of Forgetting as an important skill in research; the ability to let wrong shapes fade away.)
While we're on the topic, I'll note that Logan-concepts, rare as they are, are WAY more likely (than Duncan-concepts, for example) to transfigure or hijack my thinking shapes. I'm not sure what's up with that yet.
(Something something, the compression level at which Duncan talks about things is not really the compression level at which I like to think, and so the concepts can't really make a home there and stick? Something to do with a scope mismatch? Not feeling confident on these guesses, though.)
Reading this, I remembered my usual reaction to what you call "setting the zero-point", which serves as a pretty good defense spell. (I don’t normally think of it as a defense; it's just my go-to lens that I apply to most conversations that help me care about them at all).
My reaction is to identify and name the thing that the person seems to care about that’s behind the setting the zero-point. You could call it “Name The Value” (though "value" is kind of a loaded term, imo).
(this move is also available when anyone is complaining about anything, which is super handy)
Generally, from there, there's an exploration of why the person cares about this thing in particular (really seeking to understand what it would be like to be a person who cares about such things) and then, after that, maybe I'll introduce some other values that I think are also worth considering within the original question, just to taste ‘em together.
(You usually can't skip to introducing other values until the person feels heard on their underlying value. I mean, you can, but they'll probably feel bad, and people tend to clam up when they feel bad, which makes further conversation-of-the-type-I-want-to-have more difficult).
Most importantly for me, this conversational move tends to shift us from speculation about the world (which often requires being able to recall facts and information, which I’m bad at) and into a realm that allows for in-the-moment observation: the realm of What People (Say They) Care About.
(Actually, it’s richer than just What People (Say They) Care About— it’s how people orient to the world; what people find salient; what kind of mental schemas they keep; what sort of mental complexity they operate at; etc. There’s a lot of cool stuff hidden behind what people say and how they say it.)
I like where your mind is at here, particularly that you’re gesturing at the want for vocabulary.
Further questions:
Where does vocabulary even come from? How does it get made? What’s the process of creating new words for a field? Is observation actually dependent on having relevant vocabulary? What is a new concept made of?
What if you want to make progress in a new field that has no vocab yet? (How do you even know there's a place to explore if no vocab exists yet? How is it found?)
The link to the book in the first paragraph is broken, and it's not clear which book by Richards Heuer you're referring to - could you add the title?
Interested! Unsure how I'll use it; will need to play around with it to figure that out. But in general, I like asking questions while reading things to stay engaged and I'm very interested to see how it goes with an LLM that's loaded up with LW context.