radex.io

aboutarchive fediversetwitter

Peak

← Back to booksAmazonAudible

Thesis: Forget about natural talent — it’s bullshit. Skills are not innate, but developed through practice — lots of it. And not just any practice, but the kind that leverages the one talent all of us have: malleability of our brains.

How we practice

Naive practice. The normal way of learning things: you practice until you get to a certain acceptable level, and then you stop. You start applying the skill, and let the thing become automatic.

This is a mistake. We think that experience in applying a skill (like in a job) naturally leads to greater skill. But research shows that this is generally not the case. Teachers and doctors with 20 experience do not tend to be better than those with 5 years of experience.

Indeed, people often get worse, as our memory deteriorates with time, and skills become outdated.

To improve over time, you need the right kind of structure at work to enable deliberate practice and feedback loops.

Purposeful practice

Practice is most effective when it always keeps you just slightly beyond the edge of your current ability.

If you never push beyond your comfort zone, you will never improve.

You will hit roadblocks and get stuck. This is normal. Usually, what you need is not to try harder, but try a different approach. Slow down, find the component of the skill that’s dragging you down, and fix it.

Adaptability of the brain

Scientists used to think that human intelligence is pretty much fixed after reaching adulthood. This turned out not to be true.

Yes, the brain is most malleable when young. There are some things that aren’t possible afterwards. And the brain changes in different ways as a child than as an adult.

But the big picture is that your brain is amazingly adaptable.

Example: cab drivers in London — their hippocampus literally gets larger with the training. Another example: vision training for presbyopia — you can literally train someone to see better without any physical changes to the eye (our brains do image processing too).

Like with muscles, changes created by deliberate practice get undone over time. If you stop using a skill, you’ll settle back to the low-level “good enough” performance.

In many areas of life “good enough” is fine. But it’s important to know that you have an option to improve through deliberate practice if you choose so.

Deliberate Practice

Here’s the most efficient way of practice we know of:

The problem, of course, is that most fields don’t have established practice techniques or objective measures. (It’s also why much of the research on expertise is done on music, chess, sports).

But you can approximate by finding the best performers in your field, and learning their practice techniques. (But beware of biases in assesing who’s “best”.)

Time

To become one of the best in the world in a field takes a lot of time in solitary practice.

Gladwell’s Outliers made the “10 000 hours rule” famous. It’s misleading, because nothing says it has to be 10 000 (depends on a field), or that by doing that much you’ll surely reach the desired level of performance. But it takes a lot for sure.

The numbers are big, but think of it as a sign that we haven’t found a limit to human expertise. If you want to reach some smaller level than “best in the world”, you can do with less practice.

Feedback

Key aspect of deliberate practice is instant feedback

You need a mechanism to tell you how well you are doing. If you have a feedback loop, you can constantly improve your skill, and calibrate your mental models. Without one, you can spend decades doing something, and you won’t get any better at all.

This is easy when practicing something where there’s an objective (or well established) standard of performance, like chess, or music.

Harder example: Radiologists reviewing X-rays. Traditionally, they don’t get any feedback on whether or not their diagnosis was correct, and so they don’t improve. Solution: build a comprehensive catalog of past (non-obvious) mammograms whose correct diagnosis is known. A trainee can practice with this database and get instant feedback. (That’s huge!)

OTOH: Surgeons get plenty of feedback, and so they tend to improve considerably with time, while most other medical professionals do not.

Note: What’s the feedback mechanism for a programmer? Speed and compile/runtime errors are obvious. But what’s the signal telling me if my code is actually written well? My own eye sharpens only slightly faster than my skill. Mentors? Code reviews?

Focus on one thing at the time: Instead of thinking of your skill as a whole, consider multiple sub-skills or aspects of it, and improve them individually. If possible, this will make your feedback loops tighter and easier to measure.

Learning vs doing

Focus on practicing the actual skill you want, not on raw knowledge about a skill.

Traditional learning approaches focus on absorbing knowledge. But it’s not because that’s actually a good way of learning. It’s merely because in a lot of fields, lack of technology made practicing impossible or impractical before you got good enough to apply a kill.

(Even though the constraints changed, there’s now a strong tradition for raw knowledge: People of the book.)

Example: Future surgeons spend close to a decade learning before they actually start applying their skill.

And doctors have so many conferences and seminars. And guess what? They’re mostly useless. Interactive sessions improve the performance just a tiny bit, and purely didactic (like presentations) not at all.

There’s no tradition in medical profession of training for actual skill. There’s a mindset that if you give somone all the knowledge, they’ll know what to do. Nope.

Top Gun approach: If you can’t practice a skill directly (e.g. too dangerous/expensive), create simulations. Try to recreate conditions of the real thing as close as possible, and practice a lot in this environment (gather feedback and iterate).

Benjamin Franklin, an example how to practice when you don’t have a teacher. For practicing writing, Franklin would take articles that were very well written, try to remember their content, and then try to write his article imitating the original’s style. Then study differences, and iterate.

Mental representations

One important way in which we gain expertise through deliberate practice is by creating a lot of specialized mental representations (mental models) relevant to the domain. (See also: Smarter Faster Better, Farnam Street)

A mental representation is a structure that allows you to sidestep the normal limitations of short-term memory. Mental models allow us to hold many more ideas in our heads at the same time, and assimilate information quickly. With robust mental representations, we can easily pick up patterns and find signal in noise.

Mental models go hand in hand with practice. Practice builds better mental representations — and better mental representations make practice more efficient (because we can notice our mistakes more easily).

Example: The single best predictor of a chess player’s performance is not the number of hours spent playing chess, but the number of hours spent studying games played by grand masters — in other words, amount of time building mental models.

Expert chess players can remember most positions of pieces on a board after 5 seconds of looking at it. But they can only do it if the board makes sense. If the positions are random (not possible in a game), the advantage disappears. That’s because chess masters remember the board not as its physical representation, but at a higher level — they have chunks and patterns stored in long term memory, and they can use them to encode the board in their minds.

Same with soccer players, climbers, … — they spend a lot of time studying plays, and when confronted with a situation, they immediately know what to do — no conscious thought required.

Note: Exactly the same with programming. We don’t look at code by simulating it in our minds. Instead, we already know all the chunks and patterns, and we see the structure of code by building on top of them. And we create high-level abstractions with corresponding high-level mental representations.

Example: Writing is good example of how mental models allow us to achieve expert performance:

Knowledge telling is how novices write. You just write down thoughts that come to your mind about a topic.

Knowledge transforming is how experts write. You create a structure of what you’re trying to explain, and through refinement of that structure, you learn more yourself (you gain better mental models on the topic).

(Mental models are super important, but difficult to explain, because you need to develop… well… a good mental model of what mental models are.)

Motivation

Willpower. There’s little evidence that general willpower (the ability to “just try harder”) is a thing. There’s evidence to suggest it’s context-dependent (we can be very motivated for one thing, but not at all for another). Therefore it’s much more useful to think about motivation (or better yet: motivations).

The obvious things: take care of yourself, sleep, develop habits. Seek support from others: teams, partners, support groups.

Benjamin Franklin’s Junto — that’s a cool idea. (But don’t forget deliberate practice is still a lonely pursuit.)

Extrinsic motivators only work at first. Ultimately you have to develop intrinsic motivation for the work.

It’s not been proven, but it’s possible that doing something for a long time makes you enjoy it more. (Could also be self-selection, though.) (See also: So Good They Can’t Ignore You)

Experts

Experts who change their fields always first reach the top of the game. That’s true for athletes, scientists… but it’s not obvious for all fields. For example art: even Picasso first became accomplished as a classical painter before he developed his own style.

When you build a ladder to the top of the field, you gain a deep understanding of how to add another bar to the ladder.

And so we follow the leaders, the pathfinders, showing the whole field the way. Once on top, they expand what’s known to be possible, and they push the whole field forward.

Natural talent

Natural talent is bullshit. Seriously. Sure there are some innate differences between people, but in the end, they’re just starting points and multipliers — skill and practice trump all.

Child prodigies? There’s no compelling evidence for existance of any real prodigies. Not a child with an amazing skill that couldn’t be explained by their practice regimen.

Anti-prodigies? Again, no compelling evidence. People who can’t sing haven’t developed the skill to sing, because once labeled as “someone who can’t sing”, they wouldn’t dare to do it and practice. (Actual neurological tone deafness is a thing, but extremely rare.) In cultures where everyone sings, everyone can sing.

IQ helps at first, but then among elite performers at any field, there’s no correlation.

In chess, you can see a correlation between IQ and performance only at the very beginning. But then, as you practice, the difference dissipates. (In fact, you can sometimes see an inverse relationship. Lower IQ means you’re pushed to practice more at the beginning, and once you equalize, you’re already in a habit of practicing.)

This is true even in fields like science. There, you need to reach some minimal threshold of IQ — but after that, there’s no strong correlation.

(btw. IQ, as a universal measure of intelligence, is also bullshit.)

Big issue in believing in innate talent is that it’s a self-fulfilling prophecy. If you seem good at something, you’re encouraged to “develop the talent”, you practice a lot, and you get good. And if you’re bad at first, you’ll be discouraged to pursue the thing. The prophecy is fulfilled. (Innate talent is a paradigm — it’s self-consistent as long as you believe it’s true.)

This feedback loop also explains why innate talents are so compelling — initial tiny differences naturally get amplified into larger differences. (And even more so in competitive settings. See also: Outliers)

If there is a genetic component to developing expertise, it’s probably in our ability to maintain focus and motivation for practice. (See also: Your Brain at Work, which suggests focus is also something you can get better at with practice.)

Takeaways

← Back to booksAmazonAudible