I’ve come to appreciate my gut feeling for how things in my everyday life feel over time.
To distill the process of checking in with your actual felt sense of things into a neat information-theoretic package, I’ve recently been using a 3-bit filter: Three yes-or-no questions you can ask yourself that will give you a rough sense of where you’re at with something.
Part of the reason I like this is because there is significant nuance in how you actually interpret those 3 bits. You might just sum them up, which is how I currently work with the life satisfaction filter:
Do you like who you’re with?
Do you like what you’re doing?
Do you like who you are?
On this I currently rank a |111| = 3. Life is fantastic! But I’ve also experienced living life at a 2, and even a 1 - and I’ve noticed a qualitative jump in my baseline contentment any time I manage to flip one of these 3 bits.
But you can also interpret them similar to a conditional flow, similar to the Maybe monad. My 3-bit filter on deciding what to study next starts with a random topic R popping into my head, and then me subjecting it to the next 3 questions:
Does R appear easy to learn, given what I already know?
Does R feel useful to learn about in depth, not just interesting?
Does R still appear interesting, even if I think it’s useful?
Order matters in this interpretation. Question A needs an answer of “Yes” or “No” before you even consider Question B—which makes sense, because it’s very often the case in real life too that an answer of “No, that’s not true” closes off vast trees of questions we would otherwise have much more reason to pursue. The value of asking whether aether gets dragged along by matter has to be downgraded by how likely you think it is that aether actually exists. So really you can have 0, 10, 110, and 111 - and only 111 suggests I should actually set aside the time to study the thing in depth right now.
It’s best to stick with your gut feeling about things before you investigate things in depth. But the fun part of the conditional flow is that it suggests further questions to ask yourself a few days or weeks later that might cause your gut feeling to shift like quicksand underneath the experiential floor of conscious noticing.
If R doesn’t appear easy to learn, given what I already know, what other R′ might I learn that makes it easi_er_? And might that R′ then pass the 3-bit filter?
If R appears easy to learn but not useful, why not? How am I even defining “useful”?
If my earlier 3-bit cumulative filter for life satisfaction has any 0s in it, that’s a very likely area for improvement—but what is “useful” when I have a full 111 in that already?
Perhaps at that point I decide to cut loose and just say “useful = interesting, full stop.”
Perhaps I take a look at how stable I think that earlier 111 is, and decide that useful = makes it more likely to remain there in the future.
If R appears easy to learn, and useful, but suddenly it stops looking interesting when I contemplate how much work it would take to learn R, why is that happening?
I have no clear-cut answers for this beyond “Ah, well this is probably just a hyper interest kind of interest, then.” In which case it’s a lovely subject to watch a documentary about, or maybe even read an easy book on, but probably not worth further investment beyond that.
I emphasize that these follow-up questions are not the ones you should ask right after applying the filter. If you really can’t help yourself, email future-you reminding them to look into it, maybe with a detail or two as to what 3-bit filter you used, what input you fed it, and what tremor of thought it unleashed that you decided to punt on dealing with to them because you had things to do.
3-bit filters
I’ve come to appreciate my gut feeling for how things in my everyday life feel over time.
To distill the process of checking in with your actual felt sense of things into a neat information-theoretic package, I’ve recently been using a 3-bit filter: Three yes-or-no questions you can ask yourself that will give you a rough sense of where you’re at with something.
Part of the reason I like this is because there is significant nuance in how you actually interpret those 3 bits. You might just sum them up, which is how I currently work with the life satisfaction filter:
Do you like who you’re with?
Do you like what you’re doing?
Do you like who you are?
On this I currently rank a
|111| = 3. Life is fantastic! But I’ve also experienced living life at a2, and even a1- and I’ve noticed a qualitative jump in my baseline contentment any time I manage to flip one of these 3 bits.But you can also interpret them similar to a conditional flow, similar to the
Maybemonad. My 3-bit filter on deciding what to study next starts with a random topic R popping into my head, and then me subjecting it to the next 3 questions:Does R appear easy to learn, given what I already know?
Does R feel useful to learn about in depth, not just interesting?
Does R still appear interesting, even if I think it’s useful?
Order matters in this interpretation. Question A needs an answer of “Yes” or “No” before you even consider Question B—which makes sense, because it’s very often the case in real life too that an answer of “No, that’s not true” closes off vast trees of questions we would otherwise have much more reason to pursue. The value of asking whether aether gets dragged along by matter has to be downgraded by how likely you think it is that aether actually exists. So really you can have
0,10,110, and111- and only111suggests I should actually set aside the time to study the thing in depth right now.It’s best to stick with your gut feeling about things before you investigate things in depth. But the fun part of the conditional flow is that it suggests further questions to ask yourself a few days or weeks later that might cause your gut feeling to shift like quicksand underneath the experiential floor of conscious noticing.
If R doesn’t appear easy to learn, given what I already know, what other R′ might I learn that makes it easi_er_? And might that R′ then pass the 3-bit filter?
If R appears easy to learn but not useful, why not? How am I even defining “useful”?
If my earlier 3-bit cumulative filter for life satisfaction has any
0s in it, that’s a very likely area for improvement—but what is “useful” when I have a full111in that already?Perhaps at that point I decide to cut loose and just say “useful = interesting, full stop.”
Perhaps I take a look at how stable I think that earlier
111is, and decide that useful = makes it more likely to remain there in the future.If R appears easy to learn, and useful, but suddenly it stops looking interesting when I contemplate how much work it would take to learn R, why is that happening?
I have no clear-cut answers for this beyond “Ah, well this is probably just a hyper interest kind of interest, then.” In which case it’s a lovely subject to watch a documentary about, or maybe even read an easy book on, but probably not worth further investment beyond that.
I emphasize that these follow-up questions are not the ones you should ask right after applying the filter. If you really can’t help yourself, email future-you reminding them to look into it, maybe with a detail or two as to what 3-bit filter you used, what input you fed it, and what tremor of thought it unleashed that you decided to punt on dealing with to them because you had things to do.