Longreads + Open Thread

Longreads

• Scott Alexander has a wonderful obituary for Scott Adams, who was funny, occasionally incredibly observant, and flawed. It also feels like a note-to-self, a warning that the same traits that make someone clever and universally beloved in one decade can eventually lead them to some weird places. Like Scott, I read a lot of Adams long before my first job, which was actually pretty helpful—it meant that the reality of office work, even at a big bureaucratic company, was a pleasant surprise. One of the risks of being a fan of someone when you're young is that they'll change, and so will you, sometimes in weird and unpredictable ways. The things late-career Scott Adams liked about Scott Adams were not the ones I liked as a kid. Still, he was a funny guy who admitted that he got funny through deliberate practice and was happy to share how he'd done it. Today, neither cubicle farms nor newspaper cartoons are all that relevant, but for a while he was exactly the right guy for that place and time.
• There are lots of profiles of founders, and of companies, but fewer of specific products. Sam Knight's look at WhatsApp in The New Yorker is an excelent one. The hard part about writing such a profile is that the more ubiquitous an app is, the harder it is to think about what would be different if we didn't have it—you just don't know how many things would break if Excel stopped working or if we couldn't Google things. One fun detail from early in the piece is that Jan Koum, who launched it, kept tweaking features and changing the app's name so it would show up in the app store as a new release—there's a tradeoff between building a great app and being good at marketing an app, but the skills behind them are correlated. The piece ping-pongs between corporate history and thoughts on the broader changes in communication that a ubiquitous texting app have engendered (among these are a very effective low-level political mobilization system in India, and a text conversation between WhatsApp's manager and his wife about the name of their child, where they settled on a name but turned out to disagree on its pronunciation). WhatsApp is big enough that product decisions are not about obvious wins, but about tradeoffs: anything that makes media files a more prominent part of the experience makes WhatsApp relatively worse as a chat product in developing markets. But that's the tech industry's form of karmic balance: if you build the perfect product, your reward is that every choice you make about changing it is mostly redistributing utility between users.
• Tanner Greer has more thoughts on the WASPs, who he correctly identifies as a broader group better thought of as the Eastern Establishment. For people who like studying institutions, social relationships, power, and other such topics, it's incredibly convenient that these people 1) were the most influential faction in the most powerful country in the world, and 2) are completely fair game as far as ruthlessly accurate analysis is concerned. Someone might be offended about what you say, but it's not as if Phillips Exeter is going to dispatch a team of highly-trained assassins if you make fun of them. But they're not just a historical curiosity, or an opportunity to punch down while feeling like you're punching up. They're a template that contemporary tech elites can choose to follow, or to ignore. One thing the piece points out (current tech elites: start taking notes) is that the Great Depression both discredited these elites as effective managers and planners and wiped out the passive incomes they and their heirs lived on. When people say that all correlations go to one in a crisis, they aren't just talking about tradable assets.
• Paul Elie profiles Pope Leo XIV. It feels like there used to be more public figures whose formative experiences involved getting exotic diseases on distant continents as a precondition to being worthy of some frightfully important office job. But as it turns out, the organizations you can join in order to get this experience aren't limited to ExxonMobile and the US Army: the Catholic church will do. This piece is also a good primer on inter-Vatican politics, a topic that's usually mostly trivia but occasionally incredibly important.
• John Psmith reviews George Pólya's How To Solve It. That book is weirdly durable both as a book and as a recommendation (I picked up a copy in 2006, after reading a similarly laudatory blog post that has long since vanished). But that's math for you: if you find the right way to formalize something, you'll be associated with it forever, and that turns out to extend to formalizing what it is that people do when they do math. The book is basically a list of heuristics and some connective tissue helping you figure out when to use them, and since its focus is on the process of solving problems rather than an approach to solving any one category of problems. So it's technically a self-help book.
• An important question on Read Haus: "What's an effective method for a career writer to gain experiences outside of writing to avoid their writing becoming stale?" There's a risk that people run when they started doing something other than writing, and then started writing about that topic instead: over time, they get better and better at writing about how the industry worked longer and longer ago. (If you read Michael Lewis' later books, everything is framed in reference to the Salomon trading floor in the 80s. Whereas a modern Lewis who had just quit a finance job to write full-time would be much more likely to assume more automation. Also, trading floors are sometimes the quietest place, because if everyone's entering orders electronically, or if trading consists of manually tweaking parameters for some automated system, it's quite rude to have a loud conversation there. This depends on the product being traded, though.) Anyway, the main solution for this is that professional writers should spend an otherwise-irrationally small share of their time writing, and to try to stay actively involved in business. In any one year, committing time to venture investing, public markets, consulting, and recruiting is bad for The Diff's revenue compared to spending all of that time either writing or promoting the newsletter. But over longer stretches, it's the only way for that writing to stay relevant.
• In this week's Capital Gains, we look at the background assumptions investors have, some of which stick around so long that we forget their original justification. Things like the natural correlation between equities and treasuries, or which assets are defensive, or which companies handle an inflationary environment better, sometimes include zombie assumptions that people bet on long after they're no longer valid.

Books

The Philosopher in the Valley: Alex Karp, Palantir, and the Rise of the Surveillance State: One way to understand companies is to look at what kinds of problems the people in charge get a kick out of solving. Gates enjoyed getting the best possible performance out of minimal hardware, and for a while that's what Microsoft was all about; Benioff clearly likes persuading people of things (especially persuading them that Benioff is a really swell guy), so he runs a company where sales is what it's all about; hedge funds and prop trading firms are run by people who find measured risk-taking an energizing pursuit; etc. In Palantir's case, Alex Karp is a philosophy PhD who likes exploring tricky moral arguments, and has helpfully designed a company so it presents one moral conundrum after another.

People talk about companies like Palantir being a threat to democracy because they enable mass surveillance, but the Karp-friendly approach is to say that it's actually a fantastic thought experiment to discuss in your dorm room while passing a bong around (Karp makes a few references to weed use throughout the book): would you rather have a computer watching you 100% of the time and only flagging things for human review if there were evidence of something suspicious going on, or for you to be spied on by a human observer for a random 1% of your time? Which one's actually creepier? Which one actually has worse consequences for society? And can we even treat this as a serious premise, or are there technical and social limitations that mean that you can't have the panopticon-with-nobody-looking?

These are actually fun questions to argue about. But whatever answer you get to is going to make someone think you're deeply evil. That's just something you have to deal with if you're going to build tools that governments use to carry out their functions, because modern democratic governments reflect the will of whatever portion of the population voted for the winner, and, in an efficient market in political coalitions, that means there are plenty of people on the losing side of highly salient issues.

In fact, to the extent that Palantir is a threat to democracy, it comes from another angle: The Diff wrote long ago about compromise through presumed incompetence, i.e. breaking a political logjam by letting one side create a policy they like but that they have no hope of enforcing. If that compromise roughly reflects the popular will, then making the government more efficient in that respect actually means thwarting the popular will. (Palantir usually does this in right-coded ways, but the leftier version of it would be selling software that makes it easier for the IRS to identify spurious deductions made by small business owners. People who run small businesses skew very heavily to the GOP, and have a fair amount of discretion about how they allocate expenses between personal and business use. In effect, we've agreed that the tax code should slightly subsidize these businesses relative to big companies that can't get away with the same things.)

Karp apparently has a strong enough preference for novelty that it isn't enough to run a highly productive gedankenexperiment factory. He has to add other kinds of weirdness, too. At one point, the book describes him interviewing a candidate for a job. The candidate happened to be the grandson of British fascist Oswald Mosley (not "fascist" in the sense of "right-wing and I don't like them," but fascist in the sense of "founded the British Union of Fascists"). The interview opened with Karp reciting a speech Oswald Mosley gave opposing Britain's participation in the Second World War. "[H]e went on for a few minutes, reproducing the speech from memory... When Karp finished, he executed a few tai chi moves and walked out of the room without saying goodbye." (The other Mosley was hired.)

Palantir sometimes gets accused of being a cult. When I briefly lived in the Bay Area a decade ago, it seemed that no houseparty was too small not to have a Palantirian wearing a "Save the Shire" t-shirt. The Diff referred to the cult-like aspects of the company as one of its competitive advantages. And it turns out that Karp is quite aware of this, and, later in the book, laments that Palantir is less of a cult than it seems.

Every company that was founded fairly recently and is now huge is going to be weird. If they were normal, there would be plenty of substitutes for them and they wouldn't have grown so quickly. But Palantir seems more like it's optimized to keep leaning into weirdness, in the same way that Amazon (disclosure: long) optimizes for launching tiny new projects that have the potential to grow into something that can absorb billions of dollars in capital. That means the company actually serves an important, irreplaceable social function: technological progress always creates tricky moral questions, and Palantir has designed itself to systematically raise those questions early.

Open Thread

• Drop in any links or comments of interest to Diff readers.
• What's a profile of a software or hardware product that needs to get written? The WhatsApp piece was good, and there have been good profiles of Excel, Instagram, PowerPoint, and the command line. What's missing?

Diff Jobs

Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:

• A leading AI transformation & PE investment firm (think private equity meets Palantir) that’s been focused on investing in and transforming businesses with AI long before ChatGPT (100+ successful portfolio company AI transformations since 2019) is hiring experienced forward deployed AI engineers to design, implement, test, and maintain cutting edge AI products that solve complex problems in a variety of sector areas. If you have 3+ years of experience across the development lifecycle and enjoy working with clients to solve concrete problems please reach out. Experience managing engineering teams is a plus. (Remote)
• A Founders Fund backed, decentralized AI infrastructure company that’s solving the compute shortage by offering hardware procurement, financing, software, and customer demand that allows any data center operator to become an AI factory/turn data center infrastructure into cash flow is looking for a capital markets associate to help facilitate GPU financings and support the company’s growth into an institutional private credit platform. You’d be helping underwrite a multi-hundred-million-dollar loan pipeline and encouraged/expected to find and pursue new angles for corporate development. Investment banking, private equity, capital markets, or structured finance experience within asset-heavy industries (data center/GPU, energy, etc.) preferred.
• A frontier investment firm is looking for someone with exceptional judgement and energy to produce a constant feed of interesting humans who should be on their radar. This person should find themselves in communities of brilliant people hacking on technologies (e.g. post-quantum cryptography, optical computing, frontier open source AI etc.) that are still well outside the technological Overton window. You will be responsible for identifying the 50–100 people globally who are obsessed with these nascent categories before they are on-market, then facilitating the high-bandwidth IRL environments (dinners, retreats, small meetups) that turn those connections into a community. (Austin, NYC, SF)
• A well-funded, Series C startup building the platform and agent primitives to drive operational transformation at large, complex institutions (starting with higher education) is hiring platform engineers. The work spans distributed systems, applied AI, and full-stack infrastructure, focused on deploying reliable agents that meaningfully bend institutional cost curves. (Remote)
• A blockchain company that’s building solutions at the physical limits of distributed systems and driving 10x performance improvements over other widely adopted L1s is looking for an engineer with expertise in systems programming languages (e.g C, C++, Rust, etc.) HFT, embedded systems, high performance computing or operating systems experience a plus. (SF)

Even if you don't see an exact match for your skills and interests right now, we're happy to talk early so we can let you know if a good opportunity comes up.

If you’re at a company that's looking for talent, we should talk! Diff Jobs works with companies across fintech, hard tech, consumer software, enterprise software, and other areas—any company where finding unusually effective people is a top priority.