February 10th, 2021
- This is a really excellent take on death as just one of many limitations, not fundamentally different than only being able to be in one place at one time. Resonates a lot with how I want to think of things.
I asked Catherio to post this publicly so I could share it! - the "better horse" fallacy - the apocryphal Henry Ford quote
- you want to ask your customers for lots of feedback, but first you want to get them out of their current local optimum / attractor basin
- once you shift from horses to cars, people can give useful feedback about improving the cars
- talking with Agrippa Kellum & Raj Thimmiah February 10th, 2021 re Designing a better browser\workflow for Twitter & more / tinder for tabs
- R > my notes on this idea
- Pain point with knowledge mgmt
- Manage Overload
- (always too much stuff)
- Get Long-term Value
- (easy to think)
- multi-armed bandit problem
- example tabs
- twitter, LW, supermemo.guru, roam, hpmor
- can
- assign priorities
- update those over time
- read only part of something and save the rest for later
- stuff starts with random priority, could reassign as stuff shows up
- (already getting a read that these guys are on a totally different wavelength to me)
- like they're more rationalisty vs postrationalisty
- on twitter, threads are very ephemeral
- by adding it to the queue, you could see it in months/years
- need some sort of resolution system, after finishing
- spaced repetition
- you and your research - Hamming
- takes a long time to summarize
- (maybe these guys want to uhhh be a development arm of what Andy Matuschak's doing?)
- M > OHHH you're building the process of creating the cloze deletions into the spaced repetition
- R > supermemo basically already does this, has this workflow
- can have a thousand articles and not go insane
- but the interface, a lot of people don't like
- and learning curve is really steep
- lots of friction - you have to import the articles manually
- but with our solution, you'd do all that directly in the browser
- Mal > I like some bits here, but it's not quite seeing the space the same way I am
- (maybe these guys want to uhhh be a development arm of what Andy Matuschak's doing?)
- A > he's busy
- M > but he needs people to implement stuff
- A > gonna connect stuff
- a tab you can think of as an intention, or a thing I might do
- we don't want to track that all in our heads
- browser gives us one way of storing that
- big list, not particularly ordered
- tinder for tabs lets you view that as a queue, manipulate it quickly, garden it
- tools to prune & maintain
- spaced repetition contribution is like, if we have a cool thing for ingesting a queue, but what actually do we want the datastructure to be
- with a queue, it's better than an ungardened list, but a lot of times when we engage with stuff, the optimal schedule isn't to engage with it now or never, but repeatedly over time
- particularly with memorization or learning
- but also a lot of intentions you might have - a queue is not the right metaphor
- intentions commonly have a repeating structure
- R > have the default unit of acting on something be "do the entire thing", that's terrible
- incrementally write out something in the thread
- can kind of half-ass it, do as much as you want then move on
- third time you look at something is going to be much different
- memory consolidation
- A > I want to interrupt you - you're making a persuasive case about something we haven't heard Malcolm's reaction to
- Agrippa Kellum > our contribution is focused on the juiciness algorithm
- priority
- M > 1-100 priority rating doesn't matter! it's meaningless! juiciness is about what I'm vibing on
- A > crux here
- A > if I spend a day excited about something, there's almost nothing that will reignite that
- I want to program my future self, and I can't. I'm just a slave to whatever's recent
- I'll return eventually somehow to something I was thinking about, and I'll be like "how did I ever let go of this? this was so important"
- I can't keep track of all of the intentions I have and how much they matter to me
- and because of that I have to execute a really shitty algorithm instead
- M > see I'm trusting my own sense of what matters, in each moment
- much shorter time horizon
- wanting to reply more immediately
- M > metaphor for the tinder thing
- M > filter by urls, eg all tweets
- R > explore vs exploit
- how to learn design
- M > Trinley Goldenberg interview with me
- human factors
- user never makes mistake - they do what makes sense given what you gave them
- Agrippa Kellum's notes
- Malcolm spends a lot of time excited about something and wants to follow that. I spend most of my time not excited and then I reconnect with something i was previously excited about, get excited, and scream.
- i want incrementalism as A TOOL FOR MAXIMIZING MY EXPERIENCE OF EXCITEMENT!!! like thats why i want it. so this conversation of the priority queue being *at odds* with excitement is super interesting and making me reframe a lot.
- I notice im finding it really pleasant to write notes / chat contributions without speaking while Raj and Malcolm converse.
- and I'm noticing that it's also not nearly as draining / i'm no longer feeling like I must exit soon.
- R > FOMO
- M >
- don't miss out on what is happening because fixated on fear of missing out on other stuff
- you're always missing stuff
- there's an epic party happening somewhere that you're not at
- burning man thing
- how do I want to spend this week?
- not
- do I want to miss burning man?
- because even if I go to burning man, I'm still missing 99% of burning man
- A >
- Very exciting InterRoam concept!
- > Dallas: "Well, it *is* slimy. That's...kind of the point of lubrication, is to be slimy, so that things that might otherwise snag on each other don't, and instead glide past each other."
- related to defensiveness = protectiveness & frame battles
