# Learn Exponentially

# by Petros Christodoulou • 4 minute read • #3 on Hacker News

Your long-term ability is determined by how fast you learn.

Learn slow and you won’t reach your potential. Learn fast and you might. Learn exponentially and you’ll achieve more than anyone thought you could.

How can you learn exponentially?

Self-Reinforcing Loops

Exponential growth is very rare and usually only temporary. Most things grow logarithmically or linearly instead.

Growing logarithmically means each day we grow less than the day before. Growing linearly means each day we grow the same amount. Growing exponentially means each day we grow more than the day before.

When exponential growth does occur it’s caused by a self-reinforcing loop.

For example, the world’s population has grown exponentially at times because of this self-reinforcing loop: more people leads to more pregnancies which leads to more people

Is there a self-reinforcing loop we can use when learning?

Self-Reinforcing Learning Loops

For some adults, reading is a self-reinforcing learning loop. The more you learn from reading, the better you become at reading and so the more you learn from reading.

This is however a *very* slow process. One study implies that in the best case it takes adults 10 years of reading 1 hour a day to get twice as effective at reading. Even if this is technically learning exponentially, the improvement rate is so slow that the process is indistinguishable from a linear one.

There are other ways you can learn exponentially but almost all are similar to reading in their long-term growth rate. We need something faster.

Learn exponentially

As explained in Learning is Remembering, we generally forget ~90% of what we learn within a month and we’d be far more intelligent if we forget less. The less we forget, the less we will be constrained by the limits of our working memory and the more intelligent, creative, and effective we will be.

One way of remembering more is by using spaced repetition. Spaced repetition stops you forgetting things by systematically reminding you of them over time at the most efficient moments.

Using spaced repetition to remember some information you’ve learned might look like this:

On Day 1 you are reminded of the information which lets you remember it for +1 day. Being reminded again on Day 2 means you remember it for +2 days until Day 4. Being reminded again on Day 4 gives you +5 days and then the reminder on Day 9 gives +12 days. By the 12th iteration, a reminder will let you remember the information for +24 years.

Each reminder lets you remember the information for longer than the previous reminder so this is an exponential process. It’s known as the Spacing Effect and has a large evidence base going back to 1885.

Ok, spaced repetition is an exponential process but technically so was reading, which one’s faster?

We said in the best case it takes adults 10 years to get twice as effective reading. With spaced repetition it takes only days for your time to get twice as effective. These growth rates are completely different.

To visualise what this means we can estimate (as fairly we can) how much you learn from 60 minutes reading per day vs. 30 minutes reading + 30 minutes spaced repetition per day:

The difference is extraordinary even when looking only 5 years ahead. Use spaced repetition over a lifetime and you’ll be hundreds of times more knowledgeable.

Hmm... can you really estimate learning speed like that?

The assumptions used to estimate the difference are fair and you can look at them here. Almost none of them matter though anyway. The only thing that matters is: reading takes *years* to double in effectiveness whereas spaced repetition takes only *days*. Over the long-run, this makes combining spaced repetition with reading far more efficient no matter how you model it.

How can this be?