Links (1)
Raging against the dying of the open web one plain hyperlink at a time.
Inspired by venerable fellow bloggers, experimenting with sporadic link posts.
Scott’s disclaimer applies, except without the part of having lots of commenters who will find stuff: I haven’t independently verified each link. On average, commenters will end up spotting evidence that around two or three of the links in each links post are wrong or misleading. I correct these as I see them, and will highlight important corrections later, but I can’t guarantee I will have caught them all by the time you read this.
Links
Ask HN: Books about people who did hard things? As a self-proclaimed connosieur of books about hard things, I still added an intimidatingly long list of new books to my list from this. Personal favorites that I have already read include Skunkworks on the glory days of Lockheed’s Skunkworks, Freedom’s Forge on the awakening from slumber of the American industrial titan in the pre-war and early years of WWII, and Liftoff on the early days of SpaceX, to be paired with the more recent Reentry.
The dawn of AI roll-ups: Pitching the idea that AI sets the stage for tech-forward private equity roll-ups, which hit a middleground between traditional private equity returns and venture returns. This jived nicely with my recent interest in vertical integration, although I personally don’t see myself working on this type of business.
One point of caution is that the PE people I know seem less enthusiastic about this idea than I am. Their skepticism boils down to change management being a major challenge if AI gets you 80% of the way but not 100% of the way to replacing a department or function.Shallow feedback hollows you out: In my words, be careful what human feedback you reinforcement learn from. Describes something I’ve observed, never been able to pinpoint a short description for, but am extremely afraid of. Hopefully having a short, catchy phrase for it will help me and others avoid this failure mode of writing in public.
Thinking about Nassim Nicholas Taleb always makes me sad. My brother handed me a copy of The Black Swan twelve years ago and it was a revelation, it blew my intellectual world open - Taleb seemed the perfect embodiment of the gentleman-scholar I’d always hoped to become. He saw through the pseudo-intellectual bullshit drowning the world and was working on the most fundamental questions, working at whatever hours he felt like, flaneuring around cities, reading classics and proving math theorems.
Now I can’t even look at his Twitter because of how sad it makes me. He seems to have become a cartoon version of himself, engaging in petty fights and finding the 500th way to express his hatred of bitcoiners or his love of squid ink pasta. There is no way he could write a masterpiece like Antifragile today, even though at age 64 he should be at the height of his powers as a writer. (If Taleb somehow finds this and reads this, I’m sorry, I know it’s a cruel thing to write, but I’ve heard people say it privately enough times I’d rather it be out in the open where it can be refuted).
Review: Reentry by Berger: If you’ve ever felt like, “man I wish I could learn more about the story of and characters besides Elon involved in SpaceX and less about… other stuff,” then you should read Reentry (and Liftoff). If you’re still unsure, then you should read this review of Reentry, and then read Reentry.
My favorite ever piece of business advice comes from a review by Charles Haywood of a book by Daymond John, the founder of FUBU. Loosely paraphrased, the advice is: “Each day, you need to do all of the things that are necessary for you to succeed.” Yes, this is tautological. That’s part of its beauty. Yes, actually figuring out what it is you need to do is left as an exercise for the reader. How could it be otherwise? But the point of this advice, the stinger if you will, is that most people don’t even attempt to follow it.
Most people will make a to-do list, do as many of the items as they can until they get tired, and then go home and go to bed. These people will never build successful companies. If you want to succeed, you need to do all of the items on your list. Some days, the list is short. Some days, the list is long. It doesn’t matter, in either case you just need to do it all, however long that takes. Then on the next day, you need to make a new list of all the things you need to do, and you need to complete every item on that list too. Repeat this process every single day of your life, or until you find a successor who is also capable of doing every item on their list, every day. If you slip up, your company will probably die. Good luck.
Why it’s so hard to build a jet engine: Long but good post on why building jet engines is so expensive and difficult. My biggest takeaways from this are that stagnation is real but that building big engineering projects has always been hard and that “real engineering” also involves a lot of iteration and fixing. On the second, the difference between engine and software engineering is not “waterfall vs. agile”, it’s just that because engines need to be so reliable, they do way more end-to-end testing that mimics in the wild conditions. Pairs nicely with Flyvbjerg and Gardner’s book on megaprojects.
Gwern’s Leaky Pipelines page:
A seeker once questioned Gwern, "What is the essence of the leaky pipeline?"
Gwern said, "A single drop lost at each step fills the ocean of failure."
And the seeker was enlightened.Gwern on Dwarkesh: truly the definition of “self-recommending”.
Tyler on Dwarkesh round 4: Lots of good discussion of Tyler’s views on AI. Personally, I think Tyler’s first-order wrong but a good corrective in the ecology of views that he’s clearly debating against in this episode. I know I’m being a bit cheeky, but what I mean is that I bet growth will be higher than Tyler predicts if we get AGI, but if you average his views and the median friend of Dwarkesh, you get a better estimate than either on their own.
Satya on Dwarkesh: Maybe unsurprisingly, Satya comes off as extremely sharp here. I was super impressed by his grasp of technical details juxtaposed with his ability to distill answers to future-looking business questions into succinct but substantive arguments. Feels like the closest peek at the Satya magic I’ve gotten.
Autonomous science: part I and part II: Mackenzie and Shelby on the technological trends and opportunities for autonomous science. Highly recommend.
Neural network potentials: near frontier and potential implications: Mackenzie on neural network potentials. Been learning more about the MD/NNP space, and this article was amazing for populating my backlog of threads to dig into.
The potential of neural network potentials: One such example of a follow-up thread from Mackenzie’s article. As an ML person who stopped his physics well before quantum, I found this opinion paper to be a good bridge for my ML brain to start tiptoe-ing into the way MD/NNP people think. In another total Compound extended universe victory, Tim now works at Orbital.
Power lies trembling: Richard Ngo on the game theory and virtue ethics of coups. Pairs well with Churchill: Walking with Destiny, which I read over the holidays and strongly recommend. Churchill is one of Ngo’s examples of a “Knight of Faith”. Reading the biography helps contextualize the courage required but also the tension between human fallibility and the importance of being perceived as infallible.
The story of BuiltWith: 1 employee, $14M+ ARR: I am not that much of a misanthrope, but I have to admit I think about how big a business is possible to build with a single employee a lot more than I think about the Roman empire. On top of this, the $1B, 1 employee business has become a big talking point in the age of AI agents, so this is an interesting case study of what the “digital transformation” version of this looks like. Sadly, the article itself isn’t that informative on the nuts & bolts of how this all works, so this is more of an existence proof than guide.
On becoming competitive when joining a new company: Frankly, I don’t enjoy Ludwig’s Twitter presence, but this is a great article and example of using anonymity effectively. It captures strategies that almost noone talks about or openly endorses but many top performers apply. Useful advice for people in such a situation, especially the following snippet.
One of the first things I quickly try to map out are who are the wizards of the company. Who are the 20% guys doing the 80% that matters? This is usually a month+ long process which is almost entirely vibes-based. To make the wizard list is fairly simple:
How do people talk about this person?
What does this person do? How close are they to the core primitives of what makes the company tick?
How is their code? Can I see it? Is it impressive?
Once I have a solid list, I try to get closer. I join the same channels, I read every message they post, I stalk their PRs, I try to map out in which spaces they move, what people they interact with etc. I'll even look up their Slack, Jira and git history to get some vague lines of the main large projects they worked on and the texture of their expertise.
The goal here is to simply understand:
how can I become more like them?
How can I absorb their knowledge, their skills and everything they have that I don't have.
How can I make them want to work with me?
Most important: How do I get close to them so I can proceed with the above?
Judge your coworkers: Another example, this time by “StaySaaSy” of using anonymity effectively to describe something many people are doing, but few will talk about, and then give advice on how to do it — judge people — more effectively. I definitely do recommend StaySaaSy if you’re interested in company dynamics, even if you don’t work on SaaS or intend to.
In my life, I've witnessed three elite salespeople at work. You won't like their secret: I kind of hate this article, but it really sucked me in with its stories of a guy who built social capital in a jail with storytelling and the author’s journey to becoming a top telemarketer. Also includes some interesting notes on historical perception of sales in the US.
“Mass production would be a shadow of what it is today if it had waited for the consumer to make up his mind.” But because of what scholars call “supply-side bias,” we regard 19th-century tycoons like Rockefeller, Carnegie, and Vanderbilt as Übermensch, while erasing the accomplishments of the legions of lowly salesmen. Why? Economists, generally insulated from the dirty realities of turning a buck by tenure and/or wealth, think of demand as a vast natural force to be harnessed, like wind or oil—a conception that fits hand in glove with the equally simplistic “great man” theory, which posits that some people (men) are just born great. Sounds nice, but things look a little less elegant to the salesmen in the trenches. They know: Demand is more like blood, and it has to be mercilessly extracted, drop by drop, by an army of sweaty little goblins who don’t eat unless they hit their quotas.
A Young Man Used AI to Build A Nuclear Fusor and Now I Must Weep: One of the many good reads from Ashlee Vance’s new publication, Core Memory. Waterloo kid uses best friend Claude to build a nuclear fusor as a hobby project. Inspiring, and an optimistic with a tinge of terrifying case study on what AI empowering humans looks like. “What will you use Claude to build?”
I must admit, though, that the thing that scared me most about HudZah was that he seemed to be living in a different technological universe than I was. If the previous generation were digital natives, HudZah was an AI native.
HudZah enjoys reading the old-fashioned way, but he now finds that he gets more out of the experience by reading alongside an AI. He puts PDFs of books into Claude or ChatGPT and then queries the books as he moves through the text. He uses Granola to listen in on meetings so that he can query an AI after the chats as well. His friend built Globe Explorer, which can instantly break down, say, the history of rockets, as if you had a professional researcher at your disposal. And, of course, HudZah has all manner of AI tools for coding and interacting with his computer via voice.
It's not that I don’t use these things. I do. It’s more that I was watching HudZah navigate his laptop with an AI fluency that felt alarming to me. He was using his computer in a much, much different way than I’d seen someone use their computer before, and it made me feel old and alarmed by the number of new tools at our disposal and how HudZah intuitively knew how to tame them.
It also excited me. Just spending a couple of hours with HudZah left me convinced that we’re on the verge of someone, somewhere creating a new type of computer with AI built into its core. I believe that laptops and PCs will give way to a more novel device rather soon.
Joe Betts Lacroix, CEO of Retro, on Ashlee’s podcast: Come for the info about what Retro’s doing, stay for the portrait of a Bay Area hardware hacker turned longevity CEO (complimentary).