Why you should work on Saturdays if you’re younger than thirty
Tilmann E. Bartsch, 25.01.2023
tl;dr
“Knowledge and Productivity are like compound interest” (Richard Hamming). Therefore the commonplace idea about learning as much as possible when you are young is true in an exponential way.
There's more to success than selection bias
Lately I have developed an interest in a particular reason for the highly varying amount of output created by different individuals: “Knowledge and Productivity are like compound interest”. A quote from Richard Hamming which he mentions in his great talk “You and Your Research“.
I believe that this sentence must be given careful consideration and therefore the goal of this post is
- to show, that it describes an important fact accurately, and
- to clarify what that means for your daily work.
Considerations like this are frequently rejected by people who argue that we try to find reasons for people's success when they were actually just lucky. A concept called selection bias. Although it exists at least to some degree this point is most often made by people trying to cover up their envy or laziness.
Assess the value of knowledge before you acquire it
For convenience, I equate knowledge with the time spent on a specific topic. No matter what you do, you always shape your brain by doing it. You can acquire the knowledge of knowing every episode of Stranger Things or the knowledge which enables you write a great research paper.
Obviously this example is supposed to show that different kinds of knowledge have different values. And yes, this valuation of course depends on the goal of the particular individual. Anyway, if you value entertainment higher than hard to obtain insights you probably stopped reading already.
Whatever knowledge is of high importance to you: you always have to decide what is the best thing to do next. Most engineers tend to choose something that feels very interesting to them.
Unfortunately the value of the knowledge created by this strategy is frequently suboptimal. The reason for this can be found in the chosen task which tends to take a lot of time while generating low-value knowledge.
An according to the upper description illustrative example is the programmer who spends his time reimplementing his own complicated framework. Such tasks may take weeks of work but the new knowledge acquired is of almost no value. If you already know how to write the program, and you already did so before then there is no good reason to rewrite it just to make it more beautiful.
Therefore, it is necessary to assess the worth of knowing something before taking the time to learn it.
Do not acquire dead end knowledge
One of the key factors of the success of any extraordinary start-up is an exorbitant amount of focus. If the start-up manages to keep its focus, every achieved progress builds on top of each other in a way which can finally allow it to overtake large entities which in contrast have to take care of a lot of things simultaneously.
The exact same thing applies to technical expertise. Knowledge that builds up on each other is worth more than the sum of its parts because it allows one to outperform competitors in a certain area.
Especially tech geeks are always in danger of learning a million things that do not build up on each other and therefore are not able to be the best in any field. Put orderly: They over and over acquire dead end knowledge.
Of course, it is of crucial importance to always be aware of new trends because it helps you decide on what to spend your time on next. But this does not imply the necessity of spending weeks or months on building something just to abandon it right after you are done.
It is much better to invest time in something you can build upon. A shining example of that is Chris Lattner. He started building the LLVM infrastructure which he accompanies until today. Building upon his knowledge he designed - among other things - Swift and MLIR. All of these projects are very highly adopted in the industry.
Now is better than later - exponentially
Carrying on, it is not only important what you learn but also when you learn it. By learning something, you are able to find a more valuable topic to engage with than you were before. Thus learning something useful now will pay off in small amounts for the rest of your life.
Let’s say you’re interested in building and training a transformer network using much more efficient tool than you currently use. Doing it instantly will enable you do machine learning experiments faster in the future and thus allows you to gain insights at a higher rate.
More formally, the growth of your total knowledge value is proportional to the value of the knowledge you currently have. This is exactly the same behaviour as encountered in interest rates which lead to exponentially growing assets. In the case of capital this encapsulates the effect of compound interest.
In exactly the same manner an individual's total amount of knowledge grows exponentially.
However, when an exponential behaviour is encountered one has to bear in mind the famous quote from Albert Allen Bartlett: "The greatest shortcoming of the human race is our inability to understand the exponential function”. Because of this the previous paragraphs should be reformulated in a way that can be better apprehended by the human brain:
A single additional working hour this week equals
- two in the week one year from now,
- four in the week two years from now, and
- eight in the week three years from now.
Three every-day questions
All of this can be summarised by considering that an individual’s knowledge
- should be valued,
- is worth more when building up on each other, and
- compounds, an effect equating compound interest.
This can be boiled down to the following questions which you should always be able to answer:
- How can I measure the value of the knowledge I will acquire today?
- What are the things I can only acquire after finishing this task?
- How can I squeeze out an hour of my schedule today to learn something I can build upon?
