Captain's Log #1
Welcome to Captain's Log by Alex MacCaw. This is the first in what I hope to be semi regular newsletters. The topics will be fairly random, just whatever is at the top of my mind at the time of writing. In this I've been quite inspired by Justin Mares's newsletters.
On writing
Let me start with a confession. I have a little disdain for people who spend more time configuring their note-taking tooling than actually writing notes. You know, the people who are constantly switching tools, configuring custom extensions, and generally procrastinating.
Recently, it occurred to me that I might be the worst culprit of all because, rather than sitting down to write, I decided I must first spend years creating the perfect note-taking tool. Well - Reflect is now a good enough state I have no excuse. Back to writing.
On Artificial Intelligence
There's a running joke between my friends that all we can talk about is AI. That may be so but it’s hard not to get philosophical when an AI is auto-completing most of your code. At this point I’d pay GitHub ungodly sums for CoPilot. (Please don’t tell them).
We’ve had some amazing advances in image, text, and audio generation this year, some of which I go into more depth into here. What I’m currently most looking forward to is GPT4 which I imagine will look somewhat like its predecessor, except accurate.
Take this scenario (which repeats many times a day):
- "I have a question"
- Google the question
- Research the top most promising links
- Find the answer and close the tabs
I expect GPT4 to replace that flow. I will simply ask it a question and it will respond accurately.
Now imagine if it had access to your notes as context… you might be able to ask some really deep and personal questions. Which city should I live in? Should I propose to my girlfriend? Should pineapple be on pizza? Far-fetched perhaps, but perhaps not.
It's hard not to look at all this AI stuff and think 'how can I make money from it'. Well let me tell you, I've already had some great learnings. I passed on investing in a few of the copy-generating companies, dismissing them as GPT3 wrappers. It turns out they are some of the fastest growing companies in history and now worth billions. Oh well.
For keeping track of AI news I recommend:
On taste companies
Taste is an interesting concept. It's ineffable and hard to define. But generally people agree when something has taste or not.
Taste is something that tends to get diluted the more people that are involved. You can always tell when a product has been designed by a committee or when its come from a single mind.
Sometimes taste doesn't matter. When you buy a commodity like gasoline you care more about the utility than the experience. You want to get in, get out, and get on with your day. But when you're spending a lot of time using a product, like an iPhone, every pixel matters. You are buying an experience. And that experience, to be pleasing, needs to have taste.
A taste company is one that sells a good experience. For that experience to be tasteful it needs to be dictated by a tiny group of people or, better still, a single person. This 'taste-master' must retain absolute control over every part of the customer experience, have high standards, and a clear cohesive vision.
Collaboration between humans is never going to get to the level as the collaboration between our hemispheres. Experiences designed by multiple humans will never feel as cohesive as an experience designed by a single individual. The more people involved, the more disjointed the end product.
Apple under Steve Jobs was a taste company. Apple under Tim Cook is not. You can see this shift in their marketing copy, their move into ads, and in the amount of time they spend ‘innovating’ on notches. What taste they still have is a vestige of Steve Jobs' - entropy is slowly eroding it away.
Often the best movies are the ones where the director has the most control, and the worst are where the studios meddle. Once Game of Thrones ran out original material they started writing by consensus and the wheels fell off the wagon.
When someone buys Reflect, what they are actually buying is my taste. There are a million different ways a notes app can go. Everyone has opinions. So the only way it has a chance of working is if it is designed from a single mind.
(Now to be clear I am not taking all the credit for Reflect. There's an amazing team behind it. What I am saying though is that ultimately everything gets my sign off.)
Not all companies need to be taste companies. In fact it’s far easier not to be one. Taste companies are, almost by definition, fragile. If the taste-master leaves then the company greatly suffers. And they are harder to scale - there’s a giant bottleneck at the top.
Given all of those problems, what’s the upside of taste companies? I'd say, personally speaking, they're a lot more fun to build. They create beautiful experiences and meaning in people's lives.
On mom & pop tech companies
In March 2020 I visited New Zealand for a vacation. My flight ended up being the last allowed to land. A week-long trip turned into a year-long migration. I had to figure out the ins and outs of living in the country.
Where do you buy things from? In the US the answer is often Amazon. But in NZ, as I quickly found out, there is no Amazon, no Apple Store, and few chains. The market is just too small. Instead, there is a cottage industry of small businesses.
Initially I found this quite frustrating. There’s no directory or aggregate view of these businesses - they're spread out over a bunch of Shopify stores. And Google doesn't really work in New Zealand - the main way of finding things is through word of mouth.
But as time went on I began to really appreciate this beauty in this model. Because these businesses are often tiny, maybe half a dozen people, everyone feels agency. Which means that the support is A class. And of course they take pride in selling a high quality product.
Which got me thinking, why don't we see more of these kinds of tech companies? Small, efficient, profitable, serving a specific niche and satisfied with that.
The best example I can think of is Tailwind. They make a healthy profit, create a delightful product I use everyday, and are but a handful of employees. Let's have more of that.
Interesting stuff on the internet
David Deutsch is one of the greatest thinkers alive. His books about the universe, quantum physics, and progress are brilliant if a little impenetrable. I have done my best to write some cliff notes on one of them.
His articles though, are are little more approachable. He's written a great one on how close are we to creating artificial intelligence. I disagree with the conclusion but it's good food for thought. Another highly recommended article of his is on Optimism, Pessimism and Cynicism.
Regarding Podcasts, my two favorite right now are Sean Carroll's Mindscape and the BBC's In our Time.
Regarding Music, I'm a little late to the game but Fred again's Boiler Room set is very good.