to-dos sound nice, but bloat fast.

---

I had plenty of drafts. But I just didn’t execute. So I’m lowering my standards to increase output.

The simple plan is to execute immediately when I have time, and to not try too hard.

It’s beautiful to just start

It’s okay to start. It’s okay to just try. You have to start somewhere. Your first try will not be the final product. It will be the basis of the final product.

Don’t try hard

Trying harder will not make you smarter. Intelligence is fairly fixed. Working memory is limited. You could think as hard as you want for as long as you want and you’re not going to have a much better thought than your first, unless you begin to work. The best way to progress is to get ideas down on paper quickly and have them interact with one another. You discover through working, not thinking.

Don’t be novel

Let’s be clear. The observations made in this post have been made time and time again. It’s part of Nike’s slogan, for Pete’s sake. That’s because the dumb solution is a solution, by definition. You’re not going to get to the smart way of doing things unless you are willing to just get it done the dumb way. Sometimes, the dumb solution is also the solution. For example, the solution for writer’s block is simply to write.

Often what is important for you to write is not something new to write. Don’t try to be novel.11unless the goal you want to accomplish is novelty, I suppose. Otherwise, the pursuit of novelty can get in the way. Instead, try to accomplish what you want or need to accomplish.

Don’t be clever

Trying to be clever all the time deadens you. Your best questions are your stupid questions. You move fast when you aren’t afraid of being wrong. Trying to appear clever can sometimes make you look clever. But in the long run it undermines even the goal of a clever appearance. The best way to appear clever is not to pursue cleverness, but to be curious and pursue knowledge.

Be repetitive

Don’t worry about being repetitive! It’s a corollary of not worrying about being novel.

Repetition helps you learn.22this is the lesson that spaced repetition learning proponents preach. So don’t be afraid to repeat! Symphonies have repetition. Poetry rhymes: the repetition of sound. And rhyming has many uses.

Just write more

Concision is overrated. Concision is important for the second draft. We are focused at the moment on how to create the first draft, which can be the hard part.

Don’t apologize in your writing.33or in life. Some people will reject you, and some people will accept you, but those who accept you will accept you more if you accept yourself. Just write.

What I’m actually doing nowadays

Here’s how I’ve started to put these ideas into practice.

Writing on loose leaf

I’ve embraced writing on loose leaf paper. One great thing is that loose leaf has no built-in chronology. I think this might be one reason Vladimir Nabokov wrote on notecard. It only gains its order with the rest of the loose leaf if and when you decide to slot it in. There’s no worry about ruining your notebook. There’s no cost to writing. You just write.

It’s also super easy to scan the loose leaf into a PDF once you’re done writing, especially if you only write on one side of the paper. A notebook does not have this advantage.

Free recall

Free recall is just writing on a topic without looking anything up. It’s basically returning to the days of a high school in-class essay, except there is absolutely no pressure!

Choose whatever you want to write about, and just write! You don’t have to be correct. Mistakes are part of the process. If you forget something, try to approach from a different angle. I often find that there’s a way to remember what I wanted to discuss, and the way is enlightening.

An example: recalling a spring

For example, I recently “rederived” Hamilton’s equations by thinking how Galileo’s ramp experiment hinted at the idea that something—which we now call energy—was conserved when we drop a ball. (What’s conserved here is some linear combination of height and velocity squared—the former is understood as gravitational potential energy and the latter as kinetic energy.) But my goal wasn’t to derive Hamilton’s equations or to think about Galileo’s ramps. I wanted to derive the Hamiltonian of a spring purely starting from the basic idea of what a spring was. What would cause someone to generate the notion of a Hamiltonian? Through the process of writing, I realized (which of course sounds obvious in hindsight) that you really need to develop a notion of energy conservation before you even care about energy or Hamiltonians in the first place. As soon as you start thinking about for example stopping rolling balls with springs (at which point the kinetic energy is converted into a spring potential energy) you can easily figure out what the conserved quantity is (it’s the displacement from equilibrium position squared, for a spring perfectly following Hooke’s law) and equate that to the spring potential energy. Then by inspection of the difference between the energy expression (the Hamiltonian $H$) and the laws of motion for a few different scenarios one easily can conjecture Hamilton’s equations

$$\begin{align} \frac{dq_i}{dt} &= \frac{\partial H }{\partial p_i}\\ \frac{dp_i}{dt} &= -\frac{\partial H }{\partial q_i}\\ \end{align}$$

where $q$ is position and $p$ is momentum. Note that the second equation is essentially Newton’s second law. (I still need to work through the intuition of the first equation!)

This is all elementary, but that’s the point. You can’t be afraid of the elementary if you want to understand something complex.44In my case, I am working through the derivation of the free boson conformal field theory as the continuum limit of the critical harmonic chain, which is composed of interconnected springs. But to understand that, it is useful to remember what a spring is.

Is free recall better than flash cards?

I actually think flash cards undermine some of the principles of spaced repetition learning. You might learn to remember something when prompted in a specific way but you often don’t learn to remember things out of the blue. The latter I find is more useful because the entire goal is to remember something useful when there is a non-obvious connection lurking. Furthermore, I find free recall much more open-ended and fun. Which makes it more likely I will do it again, and integrated volume beats increased structure in my mind.

Less to-do, more do

The other important thing is that to-do lists suck because you get the sensation that delaying work is actually work itself. It’s not. Todos are by definition procrastination, which is only useful in small doses. I have made the mistake several times of building beautiful to-do lists and doing nearly none of the tasks. It’s very wasteful of my time.

A Few More Tips

Start with your first idea

Usually the first idea that comes to mind suffices as a starting point.

Don’t wait for the second idea. Usually it comes after starting up the first idea.

Outline quickly

I wrote a quick outline on loose leaf before writing this. Essentially, the outline is the first draft. Or maybe the zeroth draft. Getting it down on paper helped me write much faster.

Revise super-quickly

After you write something, you’ll find it easy to revise. So revise with glee.

LLMs are good for quick critique

I find that large language models are good at quick critiques, after you have written the first draft. Their main benefit over people is that (a) they are forced to read your work and (b) they read it quickly. Take their advice with a grain of salt as they can regress you toward the mean.

For example, Claude’s main advice was to remove the full copy of my outline from the “Outline quickly” section and that the headers sounded like a self-help book. Point taken on the outline; I removed it.

But I’m keeping the self-help-y titles. If it comes off as self-help, well, that’s because I’m at least helping myself.

Publish quickly

The scrutiny will help you improve and the pressure to make it a little presentable will speed up development. Ironically, publishing quickly and often also reduces the overall pressure because you just get used to pushing things out, and each individual thing matters less.

Following through

So I guess I’ll publish quickly. Took less than three hours to outline and then write this. I wasn’t keeping track of when I started, so it might have been much less time.