Future of Coding

Reflection 11

Jan 5-16 2018

Self-reflection leads to madness

Eric Weinstein had a really fascinating comment at time 56:19 in this podcast that really resonated with me:

If I try to compute something that’s much larger than I am, my computer can’t handle that much larger system. That’s sort of why self-reflection leads to madness.

In the past few years, I have become quite the journaler. I really love it, particularly when I write about what’s on my mind at the current moment. However, I have not been loving the reflecting on posts of days past. This is somewhat surprising to me, but this quote makes it easier for me to understand why: there’s too much data.

The human brain works via compression. It relates what we see, hear, think, feel, etc to other structures in our brain that we already have – we only store in long-term memory those details that surprise us in the way they differ from what we were expecting. This is how we make sense of the complex world around us. We experience things, and then we compress, forgetting all but what we find important. Sleep probably helps this proccess along, which is why things often feel better in the morning: you literally remember less about what was making you upset.

I love my journaling so much because it allows me to be logical right now, in the mooment. Get my thoughts precise and inspectable. Oftentimes getting a thought out in my journal well phrased makes it more likely to disappear from my mind, because that thought “feels heard.” It feels integrated into my logical system and thus can disappear. It’s burden is lifted.

So going backwards and re-reading past thoughts and anxieties is annoying. They aren’t meant to be read. They are meant to be written.

With my self-reflection schedule, I have tried to co-opt a tool that I love, stream-of-consciousness journaling, into a way to keep myself on track, keep myself honest. I don’t know a better mechanism than this one to do that still, but I want to at the very least, give myself less time (only an hour or two) for these reflections. Summarizing, compressing. This is key.

Bret Victor deep dive

I didn’t feel pulled to do this, but did it anyways because it was the plan and I had a very real deadline on the horizon. Yet while doing it, I often had a lot of fun. Spending a whole day reading Tufte was great. So was reading Learnable Programming, and all the links that he references in it.

To be honest, I have a lot more links to read if I really wanted to finish the deep dive in the way that I started it. But I don’t have the feeling to do that right now, nor do I expect myself to have this urge after leaving Dynamicland. I’m in the mood to do more programming, less history and context.

Search for meaning

As I wrote here, I am asking a lot of deep questions about meaning and motivation. I am feeling more adrift and directionless these days. One of the ways that I’m feeling this is that I sometimes fish for soft job offers from random people and get a big emotional boost when they are offered. I have even considered taking jobs that I would’ve laughed at taking months ago, such as at McKinsey.

Here are some of the the things that sound like good next steps:

Finally, I want to acknowledge how much of my thinking here has been recently influenced by fantasy writer Brandon Sanderson. There are two main themes:

1) People need to be valued. People enjoy being of service. Give people jobs to build their confidence.

This is the conservative perspective on why jobs are great, and why we should be worried by a UBI, post-scarcity world, like in Kurt Vonnegut’s Player Piano. I have started to doubt this perspective, however, as I see how people with the means to live without work not work and not become wrecks.

2) Live for a cause, for a principle. This is a way to live with integrity that few people achieve.

This may be a silly, overly reductionist way to live. Also, it may be too early in my life to try to do it. I haven’t earned it yet through hard work and practice.

Yet while I believe the “lessons” I learn from his books, I doubt if I can trust them. This makes me want to:


I did a pretty good job yesterday and this morning, reflecting on my stay at Dynamicland so far. I’m mostly focused on helping people make programs without having to type them. That is, building the programming langauge tools at Dynamicland: big surprise!

I’d also love to help with fundrasing if possible, but I feel that might be a distraction for Luke and myself.

Note to self: take pictures and videos!

After Dyanmicland

I really like the list of things to think about / work on that I put together in my last reflection.

I imagine many of them will be harder done than said. In particular, I have been thinking and talking a lot about “concrete then abstract,” the principle behind the Monad Tutorial Fallacy. Bret alludes to this too: you have to first do things, then reflect back on what you did to see what you care about. In other words, I cannot create a plan abstractly like I am trying to do. I need more experience in my life so I can see what attracts and repels me, what my principles are.

Additionally, adding a time-limit to create a plan could be a useful exercise. I did this last week and it was both fun and productive. For one, it allowed me to rule out a lot of options that would require more research before I could choose them, such as becomming a video game designer.