How to Become Smarter

(With many weird tricks)

Hi everyone,

I often have slightly frustrating exchanges where people ask me how I did something and I don’t really have an answer other than “I don’t know, this just isn’t hard? Maybe I’m just smarter than you?”. This is obviously a bad answer, and it’s a bad answer on two fronts: It’s a jerk move, and it’s not helpful.

The reason is that it’s not helpful is that it lacks explanatory power. “How do you do the thing so well?” “I’m good at doing the thing.” is a total cop out. I thought I’d try to cop out a bit less and describe what “smarter” means in these contexts and try to explain how to get better at it.

The result will probably make you more productively confused about the subject rather than working as a detailed guide to smartness, but I do have some practical suggestions along the way.

(Also, based on the number of links in this post, apparently that’s what I’ve been doing all along. I’m not sure that was intentional, it certainly wasn’t consciously so).

What is smartness anyway?

Beats me, honestly.

I don’t think it’s really a single thing, and I’m not sure trying to pin it down as a single thing helps anyone. As best as I can tell, “smart” means “good at things we’ve labelled as requiring smarts (typically because they primarily require thinking hard)”. We consider being good at chess obviously a sign of being smart, and being good at boxing not obviously a sign of being smart, but I don’t think that’s because there is some simple underlying distinction between the traits required to be good at one and not the other. Instead, there are a million small things that add up to being good at one or both of these things.

(And then of course you get people who are good at both and compete in Chess Boxing)

This is not to say that “smart” isn’t real, it’s that it’s many real things instead of one. There are traits that lead you to be good at smart-people things, and there are traits that lead you to be good at sports-people things, and some of those are the same traits and some of them are different ones, and we call some aggregate of the traits that result in you being good at smart-people things “smartness”.

There’s a lot of debate as to whether these traits are “nature” or “nurture” - something you are born with or something you learn. This debate is both scientific and very political, and here I will offer you a simple and definitive answer to this question that entirely resolves the debate: I don’t really care.

(This is a very important answer to be able to give to hard questions. See The art of not having opinions)

I think it would be weird and implausible if literally 0% of the traits involved in smartness were nature. I think it would also be weird and implausible if 100% of the traits involved in smartness were nature. If you put a gun to my head and forced me to make a prediction I would probably say that I expected it to work like physical fitness - almost everyone without some underlying health condition or disability is capable of being really quite physically fit, but it’s going to be easier for some than others and once you try to hit the peak of human performance you’re going to be bottlenecked on underlying biological traits.

But almost none of this matters, because the actual distinction that matters is “Traits I currently know how to improve” vs “Traits that I don’t currently know how to improve”. There are a lot of the latter, and some of them are going to be intrinsic, and some of them are just going to be because you haven’t figured out how to work on them yet, and the distinction of which is which is largely irrelevant until you’ve spent more time on them.

Can you get smarter?

Yes, obviously.

People will talk to you about the stability of IQ over time as a counterargument to this, but I think that’s irrelevant for a bunch of reasons:

1. Most people neither know how or try to improve their smartness over time.

2. IQ tests measure smartness in a very particular way that I think is less of a bottleneck in practical situations than it is in tests (more on this later).

3. I kinda buy that IQ is in large part determined by desire to pass tests (cf. The Stanford Marshmallow Prison Experiment), and personalities seem to be reasonably stable over time (again, in part because most people neither know how or try to change their personalities) and people mostly don’t take tests after university so are even less likely to change this aspect of their personality.

People like to hate on IQ, but as far as I can tell it’s a pretty useful diagnostic instrument in some contexts and does seem to be measuring some real things, many of which are probably reasonable to call intelligence, but I think one should still be very cautious about using it as a proxy for smartness in practical contexts.

In the other direction, people will talk to you about growth mindset and how all you need to do to become smarter is to believe that you can be. I consider all of the science on this incredibly dubious, but also think the basic conclusion that growth mindset is good is so obviously true that I’m not that worried about it: If you don’t believe you can get smarter, you’re not going to try to get smarter, so you won’t. It’s not the mindset that helps, it’s the fact that you actually do the work.

I think it is very unlikely that most people have the ability to become super geniuses, especially if they’re not starting from a very young age, but that’s not the question. The question is “Can you get smarter?”, and if smartness is being good at doing smart-people things, I reiterate my initial answer: Yes, obviously.

Should you get smarter?

I don’t know, does that sound like a thing you want to do?

I think smarts is both over and underrated, in that being smart is rarely enough for everything, but it does often help, even in things where you wouldn’t naturally think of it as helping. I think being smarter is good to the degree that it enables you to achieve your goals better, but you should usually centre the goal rather than the smartness, getting smarter is just instrumental .

I do think asking “How can I become smarter?” is likely to be life improving for most people, because I think asking most questions about how to change or improve your life in some way is life improving for most people, in that it reveals interesting insights about your self and your life. Smartness isn’t special here though - the same is true for asking “How can I become hotter/fitter/happier/…?”. All of these are useful ways of figuring out life improvements.

Importantly, these questions are useful even if you go “Actually that’s not worth it to me” because they reveal things to you about yourself. They’re not quite a form of shadow work (cf. A Brief Introduction to Shadow Work) but they’re not far off in that they reveal to you alternate ways of being yourself that you could embrace if you chose to, and they present you with the opportunity to choose to or not as you prefer.

However I do think if you properly engage with questions like this you’ll almost always find some low hanging fruit that it would be life improving to work on.

How do you get smarter?

You get smarter by applying the following two principles:

• Being smart is being good at smart-people things.

• You get better at things that you practice.

So the way to get smarter is to pick things that you would need to be smarter to do, and to get good at them. There are a lot of details to this of course - different subjects and ways of practice work differently well - but it’s all elaborations on these principles.

Unfortunately as with all self-development, this is hard and time consuming work, and the two greatest bottlenecks are always going to be that you get bored and that you don’t have the time and energy to do it.

I can’t really advise on the time and energy front because I’ve always had the privilege of having a lot of time to sink into getting better at things - my best advice there is to try to integrate it a bit into things you’re already doing, e.g. applying some of this at work.

I can however advise how to avoid getting bored. Like I said above, it’s better to have goals that are not “be smarter” and integrate becoming smarter into those goals as it’s needed. If you want to pick goals that specifically enable becoming smarter, these should be things that:

• You want to do for their own sake, because you find them interesting or enjoyable (see Desire as a driver of growth).

• Are relatively low stakes, so if you don’t see the level of improvement you want it’s not really a big deal (see Safety as an enabler of growth).

• That you can see steady improvement on which actively benefits you as you go (see Life as an Anytime Algorithm).

• That have lots of intermediate steps that you can get better at individually (see How to do hard things)

This way you’ll be having fun and getting smarter as a side effect.

Ideally you should have at least two such goals, and try to let them feed off each other, switching between the two when you get stuck (cf. Constraints on skill growth).

My personal recommendation for topics on this is that almost everyone would benefit from getting better at communication skills and almost everyone would benefit from emotional management skills, and these two synergise very well together by combining them into doing journaling. As a bonus, people who clearly communicate and present confidently seem a lot smarter (and in some sense are, because these are both smart-people things to be good at).

So, if you want my most straightforward possible recipe for becoming smarter, you should:

1. Start a weekly newsletter or blog.

2. See a therapist and/or start a mutual support group (see Notes on Running a Mutual Support Group).

3. Do morning pages and/or another journaling habit (see Writing to Understand).

My guess is that after a year of that most people will find themselves substantially smarter than they were going in.

It is, of course, suspiciously convenient that these are the things I think most people would benefit from doing anyway, but I don’t think it’s necessarily surprising that the things I think are important are also the things I think help people do better at the things they perceive me as being good at.

What does getting smarter look like?

The two traits that I think are most obviously smart are speed of thinking and strength of memory (there are actually a whole bunch of different types of memory, and all of them are useful for smartness in their own way).

I think it’s unfortunate how much we identify smartness with these, for two reasons: They are among the hardest things to change (if I had to place bets on which smartness traits were most likely to be nature it’s these ones, though speed is definitely something that you can train for specific problem domains even if not in general), but also they are very viable to compensate for.

The traits that I think are just as important and more improvable are problem solving, and finding new and useful ways of looking at things. Both speed and memory are useful for this, but you can also definitely learn to be good at them without.

If you’ll forgive the brief philosophical digression, I’m going to talk about Searle’s Chinese Room Argument.

The Chinese room argument is roughly as follows: If strong AI, in the sense of being able to create an intelligent computer program, were possible, it would be possible to physically encode it in terms of a giant library of instructions, and a person in a room with that library could just follow the instructions to simulate having a conversation in Chinese (which the AI speaks and the person doesn’t) with someone passing letters into the room. Can the room really be said to be thinking? Does the person operating it understand Chinese?

My answer to these questions is “Yeah, sure, why not?” and “No but the person/room combination does”. This is why I don’t get invited to philosophy conferences.

But I think the more interesting consequence of the Chinese room is this: If we believe in any sort of reasonable materialist view of intelligence, then anyone can simulate someone more intelligent than them, it just might be rather slow and impractical. But that’s great, because now we have an existence proof (see Seeking out existence proofs in everyday life) that what we want is possible, we just need to figure out if it’s possible in a way that is worth it.

The Chinese room itself is aggressively impractical (it would require a library the size of a planet and a simple conversation will take years), but you just as well could just stick a human translator in the room and have a perfectly good conversation in Chinese despite neither of the two members of the room being able to hold that end of the conversation on their own (assuming it’s about something the translator isn’t well versed in).

The questions that prompted this letter largely sound like “How did you come up with those that quickly??”, but when actually trying to learn to do the same yourself it’s worth separating this out: First, learn to do it slowly and with external support, and then apply the fully general system for learning to do hard things to removing that external support and speeding up the bits that you think are too slow.

Chances are you’ll end up more reliant on external support and not as fast as people you currently consider smarter than you, and chances are that this will change the cost-benefit analyses on whether some things are worth doing, but also there will probably end up being some things that you are better at than people who are naturally smart - often when you start out bad at something and learn to get better at it, once you reach the point of being as good as people who are “naturally” good at it, you’ve acquired all the skills required for improving and you might as well keep going. (This is also an excellent point at which to start applying the Unusual Foundations approach).

What sort of external support you need varies, but I think the following are all pretty reliable:

• Even just taking more time over things is a legitimate form of support - you can arrange matters so that when you need more time you get it.

• Offload more of your thinking to external mediums (see Writing to Understand).

• Where you can, do the above in searchable forms. Using Roam or Obsidian for this might be a good call, but you could also just do it in a Google doc. This helps you use it for external memory.

• Use Anki for things you really do need to remember (I personally don’t get on with Anki at all, but some people seem to swear by it)

• Develop explicit protocols for idea generation and other tasks that require “more smartness”. e.g. the random page protocol I outlined in A Guide to Starting a Daily Writing Practice is an example of such a thing.

Will this work?

Honestly, no, probably not.

All of the things I’m suggesting are good ideas which will help most people, and I think most people will “become smarter” by following them, but I think it would be very easy to follow all of this advice and not become noticeably smarter outside of the very narrow areas you work on improving, and I’m not quite sure what to do about it.

The thing about smartness that I haven’t quite figured out how to help people to get better at is a kind of reflexive attitude to drawing connections between things. It’s the attitude I described in Thinking through the implications, reflexively applied to all the things. It’s a pervasive attitude of curiosity about thinking and ideas in general, which leads to just endless practice of the varied body of skills and abilities that one can think of as “smartness”.

This particularly pays off when it is applied to your own thinking by means of Metacognition (see Self Curiosity), because it lets you examine your own thinking and try to do it better.

And I do think this requires a kind of active interest in thinking rather than just something you do because it’s work, because this is all something that requires practice, and by adopting an attitude of being reflexively interested in thinking you will naturally fit in far more practice than you otherwise would.

I don’t think this attitude is intrinsic or immutable, but I’m not quite sure what works for cultivating it. My guess is that surrounding yourself with other people who are interested in the problem of getting smarter (see Maintaining Niche Interests) and adopting norms of self-improvement in the group (see Norms of Excellence) will get you a long way there.