How Much Faster Can You Learn?

Say you were to optimize everything you possibly could for a particular learning goal. You used the best time-management system, you picked exactly the right tasks that would drive learning improvement, you applied tremendous motivation and effort. How much faster could you learn something than the status quo would expect?

The answer, of course, will be: it depends. But that’s rather unsatisfying, if true, because the crucial question is: what does it depend on?

Asking how much improvement over the status quo or expected rate of progress is an interesting question in many domains. In some areas, there might be reasons to suggest you could learn something much faster, maybe two or three times as fast. In others, you may not be able to beat expectations much at all, and in others you might actually do worse.

In this article, I’d like to go over what I think some of those conditions are, so you can plan your own expectations.

1. Is the Status-Quo at a Competitive Limit?

Competitive pressures define many areas of life. The type of competition you face can give a good picture of whether it makes sense to be optimistic or pessimistic about your odds of success, as I’ve written before here.

Many activities for learning are actually fairly lousy compared to a well-known alternative. Classroom learning of a language, without extensive extracurricular practice compared with learning via continuous immersion, are pretty clear examples. The former is so much worse, that it’s not unreasonable to expect massive improvements over the status quo, even if you aren’t especially smart or talented.

This situation can happen because there’s no competitive pressure to weed out inferior language learning techniques. Instead, the method tends to fit whatever other constraints are available, which tend to be the limited teacher-to-student ratio in the classroom, classroom hours, the need for gradeable assignments and tests, etc..

On the other hand, consider a highly standardized test like the MCAT or GRE. In this instance, any test-taker is free to use almost any possible strategy to prepare for the exam. Stakes are high and so inefficient methods are unlikely to survive in the status-quo.

Here, the potential improvement you can get has to be on the margin of improving over the average human constraints (time-management, motivation, efficiency) rather than picking the low-hanging fruit of a novel and highly-effective studying strategy.

The level of competition being faced matters here as well. Improving over the average is a lot easier than improving over the elite. So if your goal is to get “good” at chess, that might be something you could optimize over the typical chess amateur. However, once you hit the grandmaster level, it’s unlikely that there will be a lot of optimizations you can make that other skilled players have conspicuously missed.

2. How Much Variety is There in Which Goals You Can Pursue?

One way you can “beat” the status-quo is to seek a different goal than is typically sought. If the pursuit has this kind of flexibility, you may be able to find innovations that can allow you to learn more quickly, with different patterns of trade-offs than others have considered.

The MIT Challenge was the best example of this. What enabled me to finish the project was that I was pursuing a highly unusual set of constraints—only exams and programming projects, with any other material being used on a discretionary basis to achieve those two objectives.

This project led to some interesting trade-offs: no credentials, no alumni network or extracurricular experiences. My experience was probably also strictly less than an actual MIT student academically, due to other limitations and my own pace. However, done in roughly a quarter of the typical time, there’s enough advantages here that it made sense to sacrifice on some other elements.

In many domains, the world actively discourages this kind of variety. Standardized tests and exams exist to create easy-to-make comparisons, so choosing non-standard approaches reduces the value of such approaches when the goal is to show off a well-defined set of knowledge or skills.

However, in other domains, the knowledge you acquire has value outside of generating comparisons. In those cases, switching to less conventional goals may enable different methods to beat the status quo.

3. How Much Talent and Prior Experience Do You Have?

Nothing is learned in isolation. Prior experiences inform new ones. If you accumulate a large amount of prior knowledge it can make learning a seemingly unrelated skill or subject considerably easier for you than someone else. And, although difficult to quantify, innate talents likely play a similar role at mediating advantages some might have over others.

Unfortunately these factors are rarely under your immediate control. However, the advantages given to a lifetime of experience can be beneficial, as you accumulate more skills and knowledge, that makes further improvements over the status quo more likely in domains that may leverage those skills.

4. How Much Effort is Typically Expended?

Related to the idea of competition is an understanding of what a “typical” effort is for a particular goal. By putting in more effort, even if that effort is simply applying a much more rigorous and systematic attack at the problem, you can beat the expected amount.

When a domain is susceptible to lots of effort and the expected effort is low, this alone can make a big difference.

This tends to appear more often in amateur domains or for non-elite levels of professional skills. Again, it’s hard to put in more effort than truly top competitors because that level of effort usually has been reached by enough to set a new “normal” for performance in that area.

5. How Unusual is Your Project?

Conformity bias is high, which means that the space of possible solution paths to a pursuit is typically under-explored. Copying and emulation, particularly of higher-status people, is an instinctive learning process most people use, and which tends to work fairly well most of the time.

Because the solution space to a given achievement isn’t fully explored, however, there’s the possibility of getting lucky and hitting upon novel approaches that simply haven’t caught on yet. Once again, the level of competition can give some insights into how likely this is, but it’s true to some degree in almost all pursuits.

More unusual approaches to learning something will have a higher variance of outcome than following a tried-and-true path to mastery. However, there are sometimes innovations that can come out of this that are rarely used, yet nonetheless, contribute to performance.

What Should You Expect to Achieve?

I tend to go into my projects with a much more conservative attitude than it probably seems from the outside. I usually suspect that significant improvements over the status quo probably aren’t possible, and then look for reasons why that opinion might be wrong, rather than assume the opposite.

I think this is especially true for extremely competitive domains, like you’d find in important standardized tests, chess, music or athletics. Here learning innovations likely still exist, but they are going to be more modest and harder to find.

However, many things you might want to learn don’t reach this rarefied air of competitive standards and the bar to innovation may be low enough that all it takes is a serious effort to overcome it. I’ve already mentioned languages, but tons of other subjects and skills exhibit similarly low bars which can be hurdled over with some thought and motivation.

  • Denis Murphy

    Nice article Scott. I also believe that skills are transferable no matter how unrelated the disciplines may appear to be at a surface level. I also think that exploring different types of disciplines can help us to learn new things that we can then bring to other fields. Your MIT Challenge looks very interesting!

AS SEEN IN