When learning something new, it’s often better to focus on learning what’s generally possible in an area, as opposed to learning specific procedures.
Creating a video about entropy required learning how to video edit. I started by watching a few beginning tutorials to learn the interface. But after that I made sure to not watch any particular 10-part course to learn editing—I just googled things as I went. I knew most of the things I would learn from a structured course wouldn’t benefit me. I was trying to learn to edit this video, and I didn’t have plans to edit any future videos any time soon.
As YouTube’s recommendation system learned that I was interested in video editing, it began to suggest videos about various specific video editing techniques. “How to do X in [my video editing software].” “Awesome Y Effect.” I watched these tutorials, but I didn’t bother to memorize the details of how to do anything specifically.
I knew that if I ever needed specific information about these things, I could look it up. Instead, I was exploring what’s possible.
By watching random video editing videos, I learned:
And I wouldn’t have learned about most of these techniques if I had not researched video editing in the way I did.
I was looking for raw ideas, not for specific how-to procedures.
In the video I was making, I wanted to visually emphasize the increase of entropy, and spreading of energy, due to a tower of blocks falling. I thought about how to visualize this for weeks, but nothing had come of it.
Then I remembered something from one of the videos I had watched about the special effect of someone changing the color of their shirt. (Source)
I realized then “colors of objects can be shifted, and also changed in real time”.
The result (1:41-1:43):
Though imperfect, I think it turned out pretty well given the circumstances.
Imperfect, yes, but I believe these made the video better.
A friend was lamenting that he wants to learn advanced excel features, but the online tutorials that called themselves ‘advanced’ seemed to be interrelatedly redundant.
I suggested that he shift his focus: instead of focusing on learning how to via tutorials, focus on learning what to. For example, he could find someone on YouTube that does ‘excel challenges’ and makes use of obscure-yet-useful tools that wouldn’t be found in a normal tutorial.
Alternatively, he could skim the table of contents of a book on advanced Excel.
I know I’m going to forget the specifics of little-used procedures anyway, so why learn them at all? I’ll focus on learning what tools exist instead, and google whenever I have to. (It helps to keep track of what sources are great for learning specific procedures, though.)
Maybe the point of learning calculus, for example, isn’t so much to ‘learn how to take a derivative’, but instead ‘learn that any problem involving rates requires derivatives’. The former is an easy problem to look up, the latter is not.
I think the nature of expertise could be knowing how to solve a problem in abstract, and less about necessarily knowing how to execute it.
For example, Being a good programmer is knowing when to use the right tools (algorithms, data structures, etc. etc.) at the right time—but not necessarily implementing them. This is why the best programmers don’t program much themselves— instead they direct other programmers to implement their general plan.
When you have limited time, and when details about the specific steps are just a Google away, don’t focus on learning specific steps for how to do things—
Focus instead on exploring what general tools and methods exist to solve problems.
I’m experimenting with using this idea in my spaced repetition system. Sometimes I want to add something to memory that I know might be useful in the future, but isn’t yet useful. Instead of spending the time condensing the information that might be useful, I’ve started exploring a third option: saving the source of information. [Just in time vs just in case.]
Example: Cue: “What to say when negotiating?” — Answer: “Google ‘Ramit Sethi Script for X’ ”
Example: Cue: “[Want to learn X?]” — Answer: “[Read X book]”
At least this gives me a starting point in the future, without having to spend my time on something that might not ever be useful.
Another benefit here is that this method is resilient to updates of the source.
(Another benefit: to do something I’ve never done, I’ll probably need more information in the future anyway.)
* Sometimes the table of contents is more useful than the details within? Of course if you have unlimited time and memory, the details are great too, but otherwise I think this is the most effective action for optimizing the set of all domains in one’s life.
Posted 2020 September 5, updated 2020 November 21