April 2023: Network State, Data Stacks, Ice Cream
What I Did This Month
For PostHog, published The modern data stack sucks, what is a growth engineer, and some tutorials and videos. Launched a new Substack with Luisa named 100 Dinners about our dinner hosting experiments, wrote The Dinner Party Network State. Tried tweeting more to mixed success.
WeWorked, celebrated Orthodox Easter, upgraded my living room (my books are now on a shelf, not the floor), watched soccer, went to some tech meetups, played basketball, went to a knife skills cooking class, and ate too much ice cream.
Thoughts
There are portals to other ways of living in this world. Jumping into a portal when it arrives in front of you is scary, but it’s also one of the most deeply human aspects of life. Most people don’t get many chances to jump: changing jobs, moving cities, relationships, friendships, challenges, and adventures. You can look out for them, and jump when the time is right. Don’t be afraid to stare into the abyss in search of them.
People know energy matters a lot, but still underrate it. Part of the reason startups are so disruptive is that they are better vessels for channeling energy.
Working around strangers makes you how much passion about what you're working on matters. You can hear it in some people's voices that they just don't care. They are waiting for some jolt that will never come.
LLMs are often bad at giving the correct answer, but they are always confident. People’s perception of their competency shows how much impact confidence can have on your belief in something.
ChatGPT has likely made a massive positive impact on humanity purely by making people more introspective. It helps people explore their ideas as well as question if ChatGPT’s responses are true.
Tech bros should make more political donations. “Money talks.” The number of people who vote in non-national elections is relatively low (like under 50%), think of how low the number is that donate a significant amount of money. People do lobbying because it works. Politicians are people too, small scale interactions matter.
When in doubt, write your worldview. You have some way you see the world, share that with others. Explain why they should care about your worldview. Try to convince them to join you.
Having your friends live near you provides some happiness gain equal to thousands of dollars (in income), so why don’t more people subsidize their friends living near them? If not with money, then with the equivalents of money (labour, work, food)?
To be successful on social media, you need to do one of two key things: put 10x more work into posting, or be 10x more shameless.
More important than talking to users might actually be listening to them. They can be frustrating, and many listen for what they want to hear, but hard truths remain hard truths.
The insight for stealth plane technology came near the end of a 9-year-old, 40-page, extremely dense, Russian technical paper. If you do the deep work, you get the rewards.
If you do whatever it takes, you will win. You can prepare for more situations than you realize. Prepare for the conversations you will have, places you will be, and things you must do. You can seem effortless on the first try with preparation.
Recommendations
I’m a long-time Notion user. One of its best parts is that it adds features without adding complexity. This is the ideal of software, and I found two articles that explain how they manage it: one from a product angle, and another from an engineering angle.
Websites are art, it’s too bad many are so temporary and quickly lost. It is only posts like this one on the Next.js website that capture a site’s beauty for longer.
Justin Murphy writes about how modernity is filled with unseen lifestyle risks. We are living in more new ways than ever before, and many of these ways of living probably won’t work out.
You should tell people when you think of nice things about them. Also Derek Sivers on the Tim Ferriss show, he’s a thoughtful guy.
A Venkatesh Rao banger about how to create environments that enable permissionless research, which historically he argues, has led to the most impactful science.
Instead of looking at life as a grand path towards some meaningful destination, it might be more valuable to look at it non-linearly.
Early computing pioneers had a much deeper vision of how humans and machines could work together than you realize. What you’re seeing now with LLM-powered tools was a long time coming.
Tech talent VORP era. Important secondary literature for understanding the Iliad. Maybe the best people are excellent across many domains. Be suspicious of advice that separates you from great things in life. Hedonic adaption is a lie. AI is a shift, but you still have to build a good product. Status games are important. Canada’s quiet decline.
Twitter is Calvinball. We wear a lot of plastic. The ending to Transformers goes hard (shout out to Linkin Park). Lex Fridman on the new Lego Jurassic Park set (and Minecraft). Haaland raw milk appreciator. Get ready for AI war? The violin will remain. The nose-clearing Konami code.
Upcoming
I haven’t been reading enough books lately. Will carve out more time to read books this month. Need to focus on hard, important books.
Traveling, going to LA (now back which is why this newsletter is so late). Enjoying the nicer weather as much as possible.
Writing lots.