Cringe == improvement

Cringing at your former self means that you’ve improved. You always cringe at how stupid, or inexperienced, or socially awkward you were – but are not anymore.

In this sense, you should celebrate every cringe that you have for yourself. If you find that your self from 5 years ago isn’t that cringy, that means that your learning and growing curve has stopped.

Obviously, our cringiest times are our adolescent years – that’s the time when everyone hits real lows. The time of drunken messaging, obviously bad life choices, extreme social idiocy… The list goes on and on.

I once asked myself this: “If I know that I will cringe at myself two years from now, what could the possible causes of the cringe be?”

Like, if I know today what are the things that my future self will see differently, then I should do all I can to speed the process up and just skip the present stupid me, and try to imitate the smarter future me as much as possible.

Unfortunately, I don’t think I succeeded at all in this. For example, I recently reviewed some business emails from just two years ago and found them cringy – even though I wouldn’t cringe that much at my Facebook or blog posts from that period.

Here’s the deal: the domain in which I cringed (at my former self) happened to be business – something that I wouldn’t have guessed two years ago when I asked myself the future cringe question. I mean, I was way off. I thought that I would cringe at my impatience or lack of meditation. What I really cringed at (just yesterday) was related to business.

If you’re interested what the concrete cringy thing was: I didn’t really get what contracts meant. In particular, I didn’t really understand what was the exact business relation that I was in. It’s clearly visible from my emails, I’m just lost. I don’t understand the things I can do or the things I can’t do. Like, imagine a police officer who isn’t sure if he’s allowed to arrest someone or not. Sounds silly, right? Police should obviously know such things and act with complete understanding of the situation.

I’m not really going anywhere with this – except saying that I haven’t found a way to predict what the future me will cringe at. I still think that this would be a great method… If only it worked.

Now, I don’t bear any hate towards my past self – I just find it cringy, but in a funny way. Now, I’m looking forward to finding out what I’m currently doing wrong. Unfortunately, I’m going to have to wait a couple of years to find out. But hey, at least I’ll find out!


Problem-specific and tool-specific

When you have a problem, you can try to solve it from two angles: from the angle of the problem, or, if you’re lost, from the angle of tools.

Let’s say that you don’t have money. You can analyze your situation and make the most reasonable plan that fits your situation – this is solving from the angle of the problem itself.

Or you can think of any number of tools – for example, a tool might be asking yourself what would person X do – and try to solve the money problem like that. “What would Delboy do?” – and then you try that. You just “blindly” apply the tool and see what the result could be.

This approach – solving from tools, not from problems – can be useful when you don’t know what else to do, so you just throw your entire toolbox, one tool at a time, at the problem, and see if anything new will come up. An analogy: you’re trying to crack someone’s password. Approach A: think about what that person would put as a password. Put a keylogger on their computer. Watch their keyboard when they type. Create a phishing site for them. Approach B: Just try to type a bunch of random things, see if it will stick.

One tool that’s particularly useful is asking yourself “What is the easiest solution to this problem?” Even if you have a problem-specific solution, you may still sometimes prefer to take a step back and ask yourself this question.

For example, there are many possible solutions for congestion problems in cities. To solve these problems, you could make metros, new train lines, experiment with Uber, taxi, air drones, whatever. The simplest solution would probably be to make the residents use bicycles and existing public transport much more. That’s the easiest solution. That’s the low-hanging fruit. Invest more effort into that which already works.


Changing things, especially behavior, is not just one big step, but many small steps during a long period of time.

Lose weight? People want it to be a liposuction, but it’s just picking the same few meals and sticking to them for a year, or two, or more, depending on where you’re at.

Save money? People want it to be a jackpot, a winning lottery ticket, but it’s just choosing not to eat out over, and over again. Saying no: to restaurants, to expensive clothing, to anything that’s not crucial.

By no means does this mean that we should not set ambitious goals for ourselves. Dream big, but understand how it is you’re going to get there. You won’t just hop on top of a mountain – you’ll climb it little by little.

All this is contained in the Japanese philosophy of Kaizen. Kaizen is, in essence, choosing to think about self-improvement as an everyday choice, where you consistently do things that improve something just a little bit, but given enough days, this accumulates and you have giant improvements over time. Self-improvement for patient people.

Now, one common mistake that people make is to act like everything always takes time and hard work. Sometimes, the answer truly is in working smarter, not harder. Climb a mountain? If you just want to get there as fast as possible, just rent a helicopter. Sometimes there actually are solutions right before us, but we say to ourselves: “Kaizen!” and just continue stupidly on our way. The key to wisdom is knowing when you have a helicopter available, and when you actually have to walk the mountain.

The enemy: being short-sighted. It is easy to forget that the small improvements that you make today are a significant part of a wider image, so remember that wider image. Every small thing counts.

Ranked problems

When you don’t know what to do with your life, you try to solve the most important problems.

And even if you do know what to do, you may want to consider going for the most important stuff.

Unless you have a very strong analytical toolbox, it’s hard to distinguish between the relative importance of, say, problem #3 and #4. They may seem like you could work on either one, as they are both extremely important. (You may still be able to distinguish between problem #3 and, say, #3458 – e.g. solving global poverty vs. solving pollution in a small creek in some village.)

If that’s the case, you may want to consider thinking about problems in tiers, instead of particular problems. For example, problems #1 – #10 are Tier A problems: AI alignment, global nuclear war, catastrophic climate change, pandemics, global poverty etc.

In that case, it’s okay to pick any of the Tier A problems, as they all have, more or less, the same importance class. Then, you could think about picking the problem that you have the most chance do be successful at.

For example, if you’re a brilliant mathematician, you could work on AI alignment instead of global poverty, as this is a better use of your existing talent. Talented people can always “make it work” in a lot of fields, but here, you probably should be picky, and choose the thing that allows you to be most useful, considering your competitive advantage.

What if you have no competitive advantage? What if you’re fresh out of school, relatively good in a lot of things, but not a specialist in any, and without a track record or experience to point you towards a certain problem? You could:

  • filter by most important
  • filter by easiest

You could either work on what seems to be the most pressing issue even within Tier A, or you could choose the strategy of picking the lowest hanging fruit. Like, if you have a 2% shot to solve AI, but a 75% shot to solve global poverty, it’s an obvious choice. You should probably look outside Tier A if you have very large chances of solving things. (How you know what your chances are is your best informed guess – the stronger your analytical toolbox, the better the guess will be.)

And what if you have no clue whatsoever and you don’t even know how you would start in any of the most pressing problems, regardless of the important/easy filter? Then you go get skills and knowledge that can be used in a wide area, and that’s a necessary precondition for you doing something in Tier A. You go serious with self-improvement. And in this case, self-improvement is not a consequence of you doing quests – self-improvement is a quest of itself, a necessary level-up for you to even be offered the main quest lines.

In short, you should probably read this.

Sink or swim (or get a life belt)

I used to be a proponent of  the sink or swim method (or baptism by fire): put someone in a challenging situation and watch them grow.

I’ve since changed my opinion because “sink or swim” is a mental shortcut like the ones I wrote about in Maxims. I think I will compile a sequence of short posts, challenging such maxims, shortcuts, and metaphors that we all use and often live by. Let this be the first one.

The problem with sink or swim is not that it doesn’t work. It works – and I would still do it in many situations. As with all of these mental shortcuts, the devil is in the details. Sink or swim doesn’t work in certain cases, and it’s important to know in which ones, and to stop blindly following a compressed life philosophy file when what you need is obviously something else.

People can take a lot of time to sink, but they may sink in the end. Same goes for swimming. Let’s say you get promoted on your job. Much more responsibility. Much more work. It’s now hard to juggle personal commitments with work. Things start slipping. You get fat. Your psyche doesn’t like you being fat and unhealthy. Stress gets to you. All the time, you’re thinking “sink or swim!”, waiting for that magical moment when you’ll finally start swimming. It doesn’t happen. 10 years pass. You’re still not in shape, you are still unhappy, and your marriage is ruined. You still haven’t sunk, but you’re not swimming either. By the time you start swimming (if it ever happens), you’ll have gone through changes that you cannot reverse, like your ruined health or marriage.

One of the root problems was that you believed in the sink or swim method, which can be a good method, but it simply wasn’t applicable here. You should have jumped in and tried it out. But when enough evidence piles up that you’re not swimming that well, you grab a life belt, and try out calmer seas to swim in.


Warrior rationalists

(All of this is just a hunch and I may be wrong)

Descriptions that come to my mind when I think about “rationalists” (people interested in Less Wrong, Slate Star Codex, AI alignment, effective altruism etc.):

  • smart
  • high IQ
  • high achievers
  • rich (or with definite potential to get rich)
  • interested in: abstraction, math, technology, philosophy, debate
  • depressed
  • not very good at social skills (but not completely bad either)
  • white and male (mostly)
  • physically unfit

Obviously, I may be wrong, but that’s the general “vibe” I get online and offline. In short: geeky types.

To phrase it in “rationalist” terms, the set of possible experiences is usually defined in terms of the list above. Not that there is anything bad there – I think it’s amazing that there is a rationalist community and that people are learning how to think better. But there are certain qualities that I think should be cultivated – especially by the rationalist community, which I’m fond of – and these qualities usually come from experiences mostly outside of the world of rationalists.

In general, I think it’s a good thing to be very tough. I think it’s good that people learn respect, confidence and management of aggression through the practice of martial arts. I think it’s good that people move much more than 45 minutes every second day. I think it’s good for people to crawl through mud, experience physical hardship, go without food, obey commands, issue commands, get in touch with their immediate surroundings and support systems, like trying to grow food on their own, or to hunt, or to build a house, or to fix a car.

As a rule, I do not like softness and excessive intellectual “flexing”. Being geeky is alright, but I like better when it is paired with strength (and wisdom, but that’s for another post).

End notes and possible mistakes:

  • I don’t know if anything would significantly change if every rationalist was also tough and strong. I suspect it would – it seems to me that there are certain formative processes that you can undergo, but their outcomes aren’t easily measured (if at all). Martial arts, for example, are one of such formative processes.
  • I may not be sufficiently in touch with how rationalists are. Maybe they are already like this. I could find out if I’m wrong by meeting more rationalists.
  • I don’t know how much of a priority this is, precisely because of difficulty in measuring the effect of my proposed changes to the community.
  • I don’t have a lot of specific suggestions. Maybe: cultivate a practice of strenuous activity (weightlifting or wrestling or something along those lines). Get into survivalism a bit. Maybe go through basic army training.
  • It may be the case that I’m only criticizing because these are the things I do, and my subconscious thought process is “since I’m doing X, everyone should do X”.
  • This isn’t a testable prediction where I can put my money where my mouth is, but I think it’s also important to have an outlet where you can just throw the idea out there and get feedback or inspire thinking.



Being good at something feels really good, and being bad at something sucks. To be good at something, you must do the work: you must sprint the hill, repeat the notes, write out the code, and reread the books. Doing the work also sucks because you are aware at every moment how bad you are at the thing, whatever your thing may be.

When you were a kid, your life consisted of you sucking at pretty much everything and improving for years. After a couple of years, you found ways to not suck, and maybe even found certain things where not only you do not suck, you are even good at them.

Then you hit an “OK plateau” – you are good enough and you don’t invest the time and attention to further hone the skill you developed. For some skills, this may be alright. Do you really need to become the best driver in the world if that is not your profession, only your commute?

Base claim: Do what makes you happy.
Contrarian claim: Don’t try to do what makes you happy – instead, do what is useful, get good at it, and you will become happy as a consequence.
Meta-contrarian claim: Actually, if you have good reason to believe that activity X would be good for you (in terms of career, lifestyle, happiness – like, a generally good choice), then you should go for X. Just keep the contrarian claim in your mind (you may be fooling yourself, and it may be a good idea to work out alternative plans).

Since it’s not easy to become really good at things, and it also takes a lot of time, even the most intelligent and hard-working people will only ever get really, really good in a couple of areas. Some will go for a generalist skill set, some will go for a deep specialist one. I don’t know if that’s true actually. Even famous generalists in different fields (Elon Musk, Joe Rogan, Tim Ferriss, Paul Graham…) actually have deep knowledge in certain fields. This is a nuanced subject because it doesn’t really work with the binary specialist – generalist framework. There is more to it. At a higher level of resolution, we can see that these “generalists” do in fact have deep knowledge of many things. So, if we are really honest about things, when we say “generalist”, we don’t actually mean “person who knows a little about a lot of things”, we mean “person who knows a lot about a couple of things”. A specialist is a “person who knows almost all there is to know about a narrow field, and doesn’t know or do other things”.

META: In this case, it seems that the generalist – specialist framework is a false dichotomy. It’s a mental model that simplifies, but to such an extent that when you dig a little, it doesn’t really ring true or useful. We use a lot of frameworks (and metaphors) but sometimes, like in this case, they appear false and not really useful. How in general do you know that your framework/model isn’t good anymore, or isn’t good for the question at hand? What other model oversimplifies a complex reality, and how in general do you know this?

In addition to the fact that even generalists must develop at least one deep skill (sprinkles are not enough), there are also combinations. Let’s say you take a set of a 100 skills, and pick any 5 skills randomly. Most of the combinations you’ll get will not be very good, or will not really work. You’ll get things like: “cooking, math, dog training, photography, graffiti”. Not that this particular combination can’t work, but it’s distinctly different from “math, programming, writing, entrepreneurship, finance”. In both combinations, you have deep knowledge of math, but in the second combination, it seems to work really well together with other deep skills. In the first combination, it’s just there, not really connected with anything.

The point here is that random picks from a list of skills will not result in optimal combinations. And since even generalists need to develop deep skills, seeing that these things take time and effort, it’s is probably really important to craft your combination to make the most sense. You know, to kinda find your “theme”, where different things you get really good at support each other, instead of being opposed or unrelated.

When I look at my own skill tree, it doesn’t seem to work that well together. My writing and speaking skills – satisfy the really good requirement, I think. I’m a really good translator (but only for English and Croatian – for French and Swedish, I’m just okay). And I’m a good instructor as well. Parkour and martial arts are connected to each other, but not to other things. But I’m not really good at any of them, I’m just decent. I’m learning to program, and that’s completely unrelated to any of the previous skills. I also play a couple of musical instruments and cook okay, I guess. So the “theme” is something along the lines of teaching or communicating, but more could be done so that the skills support each other.

Your skills make your theme (or they don’t, if you do a lot of unrelated things). First, it’s hard to get a theme in the first place. But when you have a theme, it’s even harder to change it completely or slightly, because now, you have to get new skills, get real good at them, and your old skills could die off if you don’t maintain them. So choosing a theme is a big deal.

And finally, how deep is deep enough? You don’t have to be literally the best person ever in a skill to say that your knowledge of it is deep enough. You just have to be really good. “I fear not the man who has practiced 10000 kicks once, but I fear the man who has practiced one kick 10000 times.” 10000 is a large number, but how do we know that just 1000 is not enough? Or maybe it takes a million kicks to be real good. Or tens of millions. There is no clear guideline to this, but when you are real good at something, you probably know it intuitively.