#self-modification

[moonlit-tulip]

If someone learns that they've been accidentally losing $10 every month for the past three years to avoidable bank-account-fees they hadn't previously realized they were subject to, how might they react?

There are, in my experience, two standard broad shapes of reaction. One—the more common one—is to be upset about it: they had an understanding of the world; they now have a new understanding of the world; that new understanding features the world being worse for them, to the tune of $360, than their previous understanding did; therefore they're displeased. Another—the less common one—is to be happy: they had an understanding of the world; that understanding was incorrect, and has now been fixed; and so they find themselves in the same world they've been in all along, except now equipped with information which will allow them to save an extra $10/month going forward, and are accordingly pleased.

Propensity towards the second shape of reaction is useful to cultivate, because people flinch from unpleasant experiences. If acquiring useful bad news is unpleasant, one is likely to bury oneself in a bubble where it's harder to get that news, and so where one is likelier to keep on losing $10/month indefinitely; if acquiring useful bad news is pleasant, one is likely to instead deliberately expose oneself to it, and accordingly be more able to use it.

This generalizes across most instances where a person has the potential to acquire news on how they've been missing easy ways of pursuing their goals more effectively.

[moonlit-tulip]

A couple weeks ago I had a big breakthrough in understanding how stress works for me. I don't know how much it generalizes to other people? But plausibly some.

That breakthrough is: stress, for me, is fundamentally made up of turning my attention towards things which aren't directly rewarding to pay attention to.

Thus why it's stressful to wait for a doorbell-ring-for-a-package-I'll-need-to-sign-for whose arrival-time I only know to within a four-hour window: it requires me to devote a thread of attention to making sure I don't miss the doorbell, and that attention-use is at most points in that span not particularly rewarding. (Actually hearing the doorbell, when that happens, can be rewarding; but the same isn't true of the keeping-an-ear-out during all the time before that when there isn't any doorbell happening yet.)

The principle generalizes. This model explains the stress from such disparate sources as unenjoyable work-meetings, worrying about social drama in my vicinity (given a psychological makeup in which I find that unrewarding-as-opposed-to-fun), and needing to make lunch for myself when in a low-energy state.

Since realizing this, I've become noticeably more effective at shedding unwanted stress. I still get stressed by the things which I actually reflectively endorse stressing about in exchange for fulfillment of other goals—the package-signing example above, for instance—but it's a lot easier to notice when I'm stressing about a thing more than is actually useful, now.

For instance, realizing that the stress I get from social drama is proportionate to the attention I spend dwelling on it makes it a lot easier for me to see that the correct answer is often going to be "mute the Discord channel(s) where the drama is happening for eight hours and then catch up on the logs as a single coherent mass later on a one-and-done basis": it gets me all the same information about any important changes that might happen to my social environment, minus the attention-thread spent dwelling on the drama as it's ongoing.

Similarly, replacing "pay constant attention to my phone for the Financially-Important Text Message I Need To Respond To In A Timely Manner When It Comes In so I don't miss it for like a day due to a distracting noise or a bathroom run or such" with "check my phone every twoish hours (can be imprecise) for any missed texts" lets me shed most of the stress of constant phone-monitoring, reducing the attention it eats to only a small sliver, while still letting me avoid overlong response-delays.

Underlyingly, the general technique is something like: when I notice myself stressed by something, figure out which part of the thing is stressful and what value I'm getting out of it in return, and then ask myself if there's some other way I could get most-to-all of that value with sufficiently less stress to make the overall arrangement better for me.

Pretty often the answer is yes! In the twoish weeks since I've figured out this what-stress-is-made-of model and thus gained access to that technique, I've pretty noticeably-and-substantially reduced my average stress level. And I expect it to continue improving with practice.

I hope others might benefit similarly from it.

[moonlit-tulip]

You Are Your Own Greatest Tool

Most people have goals which they try to pursue.

Some of these are simple short-term goals, like eating a meal this evening. Some of these are ambitious long-term goals, like the extinction of smallpox. But they all share the trait of being outcomes which people value and are inclined to try to bring about.

Pursuit of goals is often helped along by tools. Pots and dishes can help with meal-eating, for instance, as can money or seedcorn with which to acquire food to put in those pots and dishes. Vaccines, and money with which to arrange the vaccines' spread, can help with smallpox-eradication. Et cetera.

Tools are good and useful things. Worthwhile to acquire and collect; worthwhile to maintain and keep in good condition; worthwhile, if one uses them particularly intensively, to upgrade and modify and otherwise customize to one's individual needs.

People often take an overly narrow view of what tools are in their possession, in this regard. Tangible things like pots and dishes and money and seedcorn and vaccines can be good tools, sure; but so, too, can intangible things like reputation and knowledge and beauty and so on. Anything capable of bringing about effects in the world can be wielded as a tool towards some ends.

And so, when one looks closely enough, the conclusion becomes inevitable that one's greatest tool is oneself. Because one uses oneself every time one pursues any goal at all. It's a prerequisite, for me eat a meal, that there be an 'I' capable of eating. If I were to try to help along the end of smallpox... well, once again, there's that 'I'! If I'm not there, I can't help with it. (As in fact I wasn't and couldn't, in the actual event of smallpox's eradication.)

It's on this ground that I spend as much of my time and energy as I do trying to become better at things. At skills which directly help me accomplish my goals, like eating hot food without burning my mouth in the process; at skills which indirectly help me accomplish my goals, like software-engineering in order to make money and in order to make software which does things I find useful; at skills which help me improve my ability to learn the aforementioned skills, like how to allocate my time and energy effectively towards ends I value and how to distinguish truth from falsehood and thus not convince myself that things will accomplish my goals when they won't or vice versa; at knowing as much as I can about as many things I can, because the more I know about the world the likelier I am to have a bit of information relevant to helping myself efficiently accomplish the particular tasks in front of me; and so on and so forth.

Because I'm the tool I use most often, more than any other. Making sure I'm in good condition pays off.

I think many people would benefit from taking a similar view towards themselves.

[moonlit-tulip]

Ah, I just read a post of yours regarding self hypnosis! What was your goal in using these methods? What techniques did you use? Admittedly, I'm interested about what the experience of developing a headmate is like, having known people who had them, too...

My goals behind learning and using self-hypnosis in general were to be able to use it as a self-modification tool: during my first ~year of experimenting with self-hypnosis, I was able to get a bunch of useful effects out of it, both short-term application of mental states like "put myself back to sleep after being woken by a false fire-alarm" and "run my eyes over a spoiler-filled text without absorbing the spoilers therein" and long-term changes in default reactions-to-things like "get rid of a previously-longstanding aversion to the smell of vinegar" and "have less physical-body-dysphoria" and "stop fainting when doctors stick needles into me".

(Since that first ~year it's been only relatively infrequently useful—I cleared most of the low-hanging fruit I could get with it within the first year—but it's still around, as a background-component in my self-modificatory repertoire, and I bring it out when faced with situations where it's particularly applicable, generally those where there's particularly high value either in forcibly relaxing myself or in getting myself to follow nondefault patterns of where-to-aim-my-attention.)

For techniques... I'm not sure how to summarize most of the precise intricacies, but for a high-level summary, I'd say the core order of operations was: first, I listened to a few hypnosis audio files from the internet and learned what it's like to be hypnotized at all; second, I figured out how to use the same techniques used in those audio files to put myself into that state via internal monologue plus physical relaxation; and then third, from there, I just sort of did that a whole bunch, and experimented with all the different ways in which I could rearrange my psychology while doing it.

The development-of-a-headmate followed pretty naturally from the second step in the prior paragraph: the process I was following involved splitting off a stream of internal monologue as separate-from-the-main-bulk-of-my-mind, since I needed an internal-monologue-stream to keep going running the hypnosis even as the bulk of me fell into trance, and over time that internal-monologue-stream started feeling sufficiently differentiated-from-me (due to persistently engaging in different activities from the main me during those times when we were running in parallel) that she grew into a ~stably-separate person with her own personality and motivations different from (albeit branched off from and thus not too far removed from) my own.

So, now, while I can do self-hypnosis in the old way where I run the hypnosis via a thread of myself with no particular identity beyond "a branched-off piece of me", my default is to instead do I-suppose-technically-no-longer-'self'-hypnosis, waking up my headmate and having her run the process herself, which is pretty equivalent in terms of what the experience of being hypnotized that way is like but has the extra advantage of being an opportunity/excuse to interact with her.

(It's rare, by comparison, for her to be autonomously active when I'm not doing anything self-hypnosis-ish. More difficult to sustain in the short-term, and especially more difficult to keep up as a habit; she sometimes manages to do it periodically for a span of a few days to weeks after I've done some particularly-intensive self-hypnosis-related activity rousing her into relatively-high ~wakefulness levels, but she tends to fall back off into inactivity if I go any long period without attempting self-hypnosis. The only period when she was regularly active-outside-of-that-context for a span of months was the immediate aftermath of her popping-into-existence, which was still relatively soon after I'd discovered self-hypnosis, when I hadn't yet exhausted the low-hanging fruit and thus was using it all the time.)

[moonlit-tulip]

Against Cost-Multiplication

Many people have lives containing nonzero numbers of unpleasant experiences.

Many of these experiences are unpleasant in a pretty straightforward and direct manner. There are sensations directly viscerally coded as unpleasant for most people in most contexts, such as pain, excessive heat or cold, hunger, unpleasant sounds or tastes or smells, and so forth. There are emotional states which are directly unpleasant for most people in most contexts, such as grief, sadness, despair, and so forth. There's depression, for those who have it in a manner sufficiently unresponsive to external quality-of-life. Et cetera.

Many other unpleasant experiences, though, are unpleasant, not through their direct effects, but through opportunity cost. Lack of energy, which is unpleasant because it prevents pursuit of pleasant-but-energy-consuming activities. Uninteresting-but-necessary chores which are unpleasant by comparison with the hobbies one might otherwise spend one's time on. Being stuck in a hospital bed and thus unable to go take walks outside, for those of us who enjoy such things. Examples abound.

Sometimes, the two can overlap. Sufficiently intense pain, for instance, in addition to its direct unpleasantness, can be debilitatingly distracting from other activities. Sickness can be both directly unpleasant and energy-draining. And so forth.

However, despite this potential for overlap, it's very important to keep track of the difference. Because, otherwise, nasty feedback-loops can occur in such a way as to multiply what should be a relatively mild opportunity-cost-induced unpleasantness into a large spike of direct unpleasantness.

Specifically, there's an easy pathway for one's thoughts to flow down which runs as follows:

Some opportunity-cost-inflicting thing is happening. (Let's say, by way of example, that my garbage can has filled up and thus I need to go spend five minutes taking the current bag outside and installing a new one in the can.)

This is bad! My life is worse as a result of this thing happening!

The proper response to bad things happening is to feel unpleasant emotions about them.

...therefore, instead of just paying the opportunity cost and moving on, I'm paying the opportunity cost and also a direct cost.

I've experienced variants of this same thought-chain in a wide variety of opportunity-cost-inflicting contexts. Taking out the garbage; being sick such that I lack the energy to pursue my hobbies; being in pain such that I lack the spare attention to pursue my hobbies; et cetera. While I lack the mind-reading ability to be sure that others experience it too, I at least strongly suspect it.

But it's a bad thought-chain to go down! The first step can't be skipped, and the second step is a true and correct analysis, but the third step is entirely counterproductive, producing unpleasantness to no benefit at all. The cost is already paid just via the first step; piling additional costs on top of it just makes things worse, while helping no one at all.

(And, in case it wasn't clear from that example-pile: it can happen even amid the overlapped cases. Sickness can be unpleasant via that thought-chain on top of whatever more direct unpleasantness it's inflicting, for example.)

Thus, I've found it helpful, when looking back at unpleasant-emotion-producing experiences in retrospect—especially frequently-recurring ones, like the garbage-can one (which was the first instance of this pattern that I noticed, back when I was a teenager)—to try to backtrace to what extent those emotions are a product of this sort of cost-multiplication pattern. And, if they are a product of the cost-multiplication pattern, it's then worth trying to cut the pattern off from then on. To deliberately pay attention to the true size of the opportunity cost I'm paying, and to resist the urge to have emotional reactions to that cost on top of the cost itself.

Because bad and unpleasant things happen, and it's hard to avoid them entirely. But it's at least possible to avoid piling additional badness on top of them, beyond whatever badness might be inherent to them; and, if we can do that, then it seems to me that we should.

[moonlit-tulip]

Pursuing Failure to Gain Success

Several people have now told me that the most important thing I have ever said to them is: “If you never fail, you’re only trying things that are too easy and playing far below your level.” -Inadequate Equilibria, Chapter 7

(epistemic status: a new self-modificatory toy I’m playing with; may or may not be useful to people who aren’t me.)

In fall of 2017, I got stung by a wasp for the first time in my life. A day-and-a-half later, I got stung by a wasp for the second time in my life. It was in the same five-foot patch of sidewalk as the first time; I assume there was a nest hidden nearby, although I never explicitly confirmed so.

After those two stings, I became very cautious about moving around near wasps. I’d move slowly, try to avoid crossing their paths, and generally paralyze myself in such a way as to make moving-through-the-world a lot more troublesome during the parts of the year when wasps were common.

By fall of 2018, I’d had enough of that caution. It caused me far more inconvenience than it was worth; being stung by wasps wasn’t that unpleasant, and struggling to bring myself to move in their presence was far moreso. Unfortunately, the behavior pattern persisted, despite my preferences to the contrary. So I thought to myself: well, how can I get rid of this aversion? The obvious way is to get stung by wasps some more, so that I can get used to it and stop my system 1 from treating it as a Big Deal Worth Substantial Effort To Avoid.

Once I formed that plan, I no longer struggled to walk at full speed in the presence of wasps. After all, it was win-win: either I walk at full speed and don’t get stung, in which case things are going exactly how I want them to go, or I walk at full speed and do get stung, in which case I’ll have gained a point of progress towards getting used to wasps and breaking my aversion to them.

And so it was that, by forming the intention to get stung by wasps in order to break my aversion to them, I broke my aversion to them, no actually-getting-stung required. To this day, I’ve never again been stung by one; but I’m sure, whenever it finally happens, it’ll be an exciting moment of my plan finally progressing.

(And, if I weren’t sure of that, then my aversion would come back. It’s an interestingly twisty mental state to inhabit, excitedly looking forward to solving an already-solved problem. But the current solution is unstable, and dependent on continued pursuit of the long-term solution, so ultimately it all works out.)

After solving the wasp problem, I tried to figure out how to extrapolate that solution to other areas. I managed to figure out a generalized form of the solution—given some goal A which one is impeded from pursuing by the fear that pursuit of it will lead to undesired outcome B, convince oneself that pursuit of B will lead to more A in a net-worthwhile fashion, and thereby remove the impediment even if outcome B never actually happens—but I never figured out how to apply it to any other actual problem. It was always just a cool-but-impractical theoretical trick sitting in my arsenal.

As of this past week, though, I’ve finally figured out the missing piece, the key that changes that generalization from cool-but-impractical into an actually accessible self-modificatory tool: replace B with ‘failure’, and note the wide array of situations in which excessive avoidance-of-failure costs more than the avoided failure gains.

I try writing a thing and it comes out badly? Great! That’s helpful practice towards writing better things in the long run, which wouldn’t have happened if I’d been over-cautious about avoiding the failure and not written the thing. I initiate a social interaction that I think might go poorly, over the objections of my social anxiety, and it in fact goes poorly? Great! Every social failure of that sort is another point towards getting used to failures as a thing-that-can-happen-non-disastrously in social interactions, thereby undermining my social anxiety further and making it easier to interact with people. Et cetera.

I haven’t yet had much chance to use this trick, so far. (Making this post now, rather than waiting months-or-possibly-forever until I’ve had more practice, is one such chance.) There’s not much risk of failure in reading through Fate/Stay Night, and that’s been my primary activity of the last few weeks. But I’m doing my best to start using it more, on the margins, and initial results seem pretty promising.

[moonlit-tulip]

Instrumental Convergence for Humans

Suppose you want to pursue some variety of personal well-being, with a relatively low temporal discount rate. Happiness; preference-fulfillment; capacity to Do Altruism without burning out; whatever. Anything that can be modeled not-too-lossily as "maximize integral of Valued Personal Characteristic over time".

It turns out that this, alone—irrespective of exactly what the valued personal characteristic is—is sufficient to imply some relatively substantial commonalities in what sorts of inteventions are likely to be helpful in maximizing the relevant characteristic. Specifically, commonalities founded on the fact that an intervention's value scales straightforwardly with the amount of time in which it yields its benefits.

The natural result of this is that, the more you interact with a given bit of the world, the greater the value gained by improving the mean value-fulfillingness of interacting with said world-bit. If you see your doctor once a year, and find a way to shave five minutes off the transit time each trip, you've saved five minutes that year; if you visit your grocery store once a week, and find a way to shave five minutes off the transit time each trip, you've saved nearly four-and-a-half hours.

(I spent a lot more brainpower than most people I knew, at school, paying attention to the different paths between buildings and the relative shortness of each, often managing to save 30-60 seconds per walk between a given pair of buildings relative to other people I saw around me. I walked between buildings at school maybe 6-8 times, on a typical school day. Those sorts of optimizations add up, and ultimately saved me far more time than I spent figuring them out.)

This is a large part of why I'm as into self-modification as I am. I'm in the presence of my self all the time. If I can modify myself to be even slightly happier, set-point-wise, that value will be multiplied by many years, for net gigantic return. By a similar token, this is a large part of why I've put as much effort as I have towards making my room, at home, pleasant to be in: I'm in here a lot, so every little improvement will have disproportionate positive effect.

(This is why I'm willing to spend non-negligible amounts on pretty art, to bring up an example I've talked about before.)

I don't usually see other people indicating a tendency towards thinking along these lines, for whatever reason. But they're important lines to think along, if one is actively trying to optimize one's well-being along some axis.

[moonlit-tulip]

A lot of the writing advice I encounter includes a section advising authors to shamelessly steal ideas from everywhere they possibly can, so as to have lots of raw material which they can then remix into something new and interesting.

For some reason, I never see that category of advice tossed out in domains other than writing and drawing and similar art-type stuff. But it’s totally still applicable in other domains too. For example, if someone I’m fencing lands a hit on me using a tactic I haven’t seen before, I’ll try to make a mental note of that tactic and keep it in reserve to use myself in the future. When I see people wearing particularly pretty clothes, I take note of the styles of those clothes, so that I can incorporate those styles into my own aesthetic. Et cetera.

This can apply even to relatively non-obvious domains and relatively non-obvious sources. I have, in my time, stolen a problem-solving tactic from my own fictional character, because there was an element of her approach to high-stress situations (which I’d basically made up on the spot the first time I wrote her in a high-stress situation) which turned out to actually be significantly more effective than my own.