Why Are We Surprised That Startups Are So Freaking Hard?, The Secret Father of Modern Computing, and More

Everything we published this week

Via Lucas Crespo.

Hello, and happy Sunday! 

First, some housekeeping: you may have noticed that we’ve been experimenting with the format of this newsletter. We’d love to know what you think of it. Like it? Miss the old one? Tell us in the comments.

In the latest episode of his podcast, How Do You Use ChatGPT?, Dan Shipper spoke to Notion AI researcher Linus Lee, who breaks down how he uses ChatGPT and Notion AI to augment his brain. Watch it on YouTube.

Now, on to everything we published this week, along with our take on the latest tech and business news.

Our stories

“Why Are We Surprised That Startups Are So Freaking Hard?” by Benn Stancil: Firefighters? Pfft. Veterans? Wusses. E-commerce founders drop-shipping knockoffs from Vietnam? Heroes. Startups are not easy to do, but also, they’re, like, not that hard either. Still, people continue to bemoan the challenges that startup life brings. Benn Stancil argues that the effort is actually the point—and the way startups win is through work. Read this for a smart, level take on how hard you should work (and how that work is necessary). 

“How David Perell Writes an Essay” by Dan Shipper/Superorganizers: You obviously like essays—that’s why you are subscribed. If you are sick of paying us for prose, this interview will teach you how to write them for yourself. Teach a man to fish, etc., etc. Read this to learn how master writing instructor David Perell writes bangers.

“How an AI Researcher Uses ChatGPT and Notion AI” by Dan Shipper/Chain of Thought: Linus Lee, an AI researcher at Notion who Dan has become friends with, has the frustrating ability to make you feel dumb—and make you like him for doing so. In this interview he shares how he prompts, cajoles, manipulates, and mansplains AI robots into doing work for him. Read this to learn how one of the best AI people in the world augments their workflow. 

“Want to Build? Technical Excellence Won’t Be Enough.” by Evan Armstrong/Napkin Math: Despite tech Twitter’s insistence, Evan strongly believes that you should read a book. It is stupid that this is a controversial stance, but alas, this is 2023. A broad humanities education is necessary because it gives you a sense of taste. Taste is necessary because if you just build cool tech stuff without any sense for what normal human beings like, you get crypto. Read this for a strong, personal argument on the why and how of humanities in tech. 

“The Secret Father of Modern Computing” by Gareth Edwards: If you want to know the future, study the past. Ed Roberts is tech’s future. He launched the world’s first personal computer company, made said company a smashing success, got burned out, and left it all behind for a field that actually provided meaning (medicine). Read this if you want to know where your startup’s great-great-granddaddy was from.  

Chain of links

LLMs solve a really hard math problem. Are language models glorified summarization machines, or can they actually advance our understanding of the world? This is a fundamental question in AI. This week it was revealed that they can make discoveries—at least in some domains. A DeepMind team used a language model to find novel solutions to a longstanding open problem in mathematics. We’ll see this kind of headline more frequently.

Progress in AI-not-kill-everyone-ism. When GPT-4 finished training, it underwent an extensive fine-tuning process that used human supervisors to help it learn how to be helpful and harmless. It worked because humans are smarter than language models. But what happens when AI becomes smarter than humans? Will supervision work then? OpenAI released a research paper saying that it could—which is significant progress.

The future of creative interfaces. Krea, an AI-first design tool, launched in open beta. It uses a combination UI of text-to-image and image-to-image tools that allows for extraordinary control in AI art creation. Expect this type of UI to become commonplace in creative tools. (Note: pair this with our recent podcast episode with Linus Lee for more on creative tools for thought in the AI world.) Dan Shipper

The napkin math

Goodnight, sweet autonomous prince. GM autonomous vehicle division Cruise cut nearly 25% of its headcount, about 900 employees, as the fallout from an unreported accident continues. Dan and I took a ride in a Cruise vehicle a few months ago and found the experience magical. Not perfect, but still, magic. Studies from Waymo show a 95% injury reduction as compared to cars driven by humans. This technology is objectively safer, but GM is kowtowing to regulators over one mistake. I can’t believe I’m about to type this, but you need an Elon-level asshole to run this kind of company, not a Detroit-based pushover. Put middle fingers up to the regulators and show that the tech works. 

Apple continues to dominate. The company hit an all-time-high valuation of $3.08 trillion. Services revenue continues to grow, the iPhone is still crushing, and the Vision Pro launch looks promising. Throw in the Fed signaling it’ll cut rates in 2024, and you’ve got a recipe for mega-moolah. Bloomberg’s headline can’t be beat: “Apple Is So Big, It's Almost Eclipsing France's Stock Market.”

Media continues to get kicked, punched, and spit upon by big tech. Most media companies (yours truly not included) rely on Google to send traffic their way. New studies are estimating a 20-40% drop in traffic for publishers. This is a catastrophic, the-business-might-die-type of result for the media. Yikes. —Evan Armstrong

The examined life

Learn from the Greek OG. You could read a bunch of self-help books, each basically regurgitating the same advice, or you could learn from the guy that all modern authors are ripping off. Marcus Aurelius is one of the founding fathers of Stoicism, a Greek philosophy that is very much in vogue. If you don’t happen to read ancient Greek, you can learn about him and his teaching in this 14-minute, 11-second video essay

“I would love to see a trillion humans.” Jeff Bezos gave his vision of the future during his first lengthy podcast interview on the Lex Fridman Podcast. Look, his biceps are big, his bank account is bigger, and everyone alive relies on his company’s services in one form or another. It’s worth listening to how someone so powerful thinks about the future. —Evan

That’s all for this week.

Like this?
Become a subscriber.

Subscribe →

Or, learn more.

Read this next:

The Sunday Digest

Reddit S-1, Gemini Picked a Stock, and My Cousin Elon

Somehow Bitcoin returned

1 Mar 3, 2024

The Sunday Digest

How to Make Yourself Into a Learning Machine, Live Fuller, Not Bigger, and More!

Everything we published this week

Oct 29, 2023

The Sunday Digest

In Praise of the Meandering Career, America’s Electrical Grid Is an Incredible Feat of Engineering, and more!

Everything we published this week

Aug 20, 2023

Chain of Thought

🎧 ChatGPT for Radical Self-betterment

Clinical psychologist Dr. Gena Gorlin’s AI-powered annual review and goal-setting session

Jan 31, 2024 by Dan Shipper

Chain of Thought

Quick Hits: New AI Features From Arc and ChatGPT

The future arrived faster than I expected

2 🔒 Feb 2, 2024 by Dan Shipper

Thanks for rating this post—join the conversation by commenting below.


You need to login before you can comment.
Don't have an account? Sign up!

Every smart person you know is reading this newsletter

Get one actionable essay a day on AI, tech, and personal development


Already a subscriber? Login