Longreads + Open Thread

Longreads

• In The New Yorker, Joshua Rothman has a thoughtful meditation on what AI does to the arts. Probably the most evenhanded look at the benefits and drawbacks of generative AI outside of software engineering. (This is not to say that the software engineers are smarter commentators than everyone else, just that it's a lot easier to see if the bug in your C++ got caught than to confirm that you've correctly interpreted a passage from Schopenhauer.) One sensation the piece captures particularly well is the realization that you've pushed a model beyond its current capabilities, and that you're actually capable of answering that question yourself. Unfortunately, in many domains it will A/B test better for the model to be confident-but-wrong.
• Kelsey Piper on "the honesty tax," or the large number of circumstances in which people are either likely to lie by accident or allowed to lie for personal benefit with impunity. It's obviously socially corrosive to split society into a group of people who are honest even when they can get away with lying, and another group of people who get admitted to better schools, offered more generous mortgages, forgiven larger PPP loans, etc., purely because they were willing to bend the truth. And it's very weird to have a system where everyone's expected to knowingly lie, and we're all supposed to develop some internal sense of which lies are bad and which ones are standard operating procedure. So not only does this lead to a less just distribution of resources, but it places a bigger cognitive burden on honest people. The post notes that there are two solutions here: one is stricter enforcement, which seems like it could claim a few prominent scalps, but the other is to just loosen the rules. For welfare in particular, this makes a great case for replacing more programs with pure cash transfers. If someone has trouble getting and holding down a job, is it really a wise policy to expect them to familiarize themselves with a complicated rulebook and keep meticulous records? We've ended up with a truly weird system where if someone is in the bottom 30% or so of earners, giving them a gift or a raise has a good chance of making them poorer, all as a result of policies whose original intent was to make them better-off.
• EigenMoomin on the strange state that is Singapore, in particular the phenomenon that Singapore is very wealthy on average, but has a system that's almost perfectly designed to limit people's ambitions. The steelman of this is that Singapore was and remains an entrepôt, with a comparative advantage in mediating economic relations between multiple big countries. If that's what history and geography gave them, it's hard for other activities to measure up. And it's still net useful work, that makes the world's logistics and financial system function a bit more smoothly, even if it creates labor alienation at the scale of an entire country.
• Maxwell Tabarrok asks: should the US have kept its empire? Specifically, should the US have withdrawn from ruling the Philippines and basically controlling the Caribbean, or kept up with that? Which sounds like a somewhat absurd question, but you can turn it around and say: in the 20th century, the US gave up some of its 19th-century territorial gains—did we happen to choose precisely the right amount, or should we have sold Alaska back, made Texas a Republic again, etc.? From a moral and pragmatic perspective, none of these are easy questions, and it's fun to think about what electoral history in the twentieth century would have looked like if a quarter of the electoral college represented the Philippines.
• Fernando Borretti on managing ADHD. (The longer this is just a lonely open tab, the more you need to read it.) Like a lot of productivity advice, this piece boils down to: Find the set of things you can realistically do, and then do all of them. Which, fortunately, has a kind of momentum; the more your to-do list is a real list of the things you actually need to do, the more instinctive it is to look at it and start checking things off; the more your time is spoken for by engagements you've actually put on your calendar, the better your calendar is as a guide to how much time you have for the to-do list. Executive function is a limited resource, and like many such resources there are high returns to spending it well.
• The Yet Another Value Blog book club is back, with a review of Railroader, the sory of the railroad CEO Hunter Harrison. Last month's book was Softwar, on Larry Ellison, and we did indeed get a lot of Softwar vibes from this one: Harrison shares Ellison's penchant for designing a product around a specific workflow and then telling customers that they all need to adapt their business to it, but Harrison was also a relentless grinder who spent much of his time on tiny optimizations rather than big-picture theorizing. Harrison is also a good answer to the question of what the people who have a natural aptitude for prop trading did before there was lots of trading—they used the same kind of mental modeling skills and quick mental reflexes to figure out which trains needed to go where.
• In this week's Capital Gains, a general theory of haggling, which turns out to be more common than you'd expect.

Books

The Story of Silver: How the White Metal Shaped America and the Modern World: It takes a lot of state capacity to have a stable currency backed by the state's power to tax—you need the ability to collect taxes and the willpower not to print money when spending needs outstrip revenue, and it's hard to have both. It's even harder if you need to borrow, because either you're borrowing in a foreign currency and eroding your own ability to repay, or you're borrowing in your own currency and convincing distant lenders that you'll maintain its purchasing power. A good stopgap for participating in the global economy has, historically, been the use of some precious metal as either the literal currency (stamping it with the face of the reigning monarch) or as backing for paper money. But this just opts into a new set of problems. If you use a single metal, then economic growth is partly tied to the supply of that metal&dmash;gold shortages like the one that occurred in 15th century Europe or in the late 19th century in the US. Use two of them, and you make monetary policy a perennial political question, and sometimes one that collides messily with other elements of the political system.

This book is a history of American silver policy that mostly focuses on the period from William Jennings Bryan's Cross of Gold speech through the Hunt brothers' attempt to corner the global silver market in the 1970s. (There's a kind of bizarre couple of chapters at the end discussing Berkshire Hathaway's silver trade, and insinuating that it, too, had manipulative intent.) As with any time you retail a broad swathe of history through the lens of a specific commodity, all sorts of easily-missed details zoom into view once you do this. For example, the US constitution sets up a legislative system that represents both the population, through the House, and states themselves, through the Senate. But states vary in size, and, in particular, some of the underpopulated ones had enough scale to merit statehood because there were valuable natural resources there, even if there wasn't much else. As a result, the US tended to have better representation for silver producers (who had lots of senators in the Western half of the country) relative to silver consumers (Boston silversmiths who would periodically complain that the government's bid for monetary silver made literally-silver silverware needlessly expensive).

A common thread in the book is that money works as an abstraction that's valued through social convention, but silver is a physical thing that exists in the ground, in bars, or in coins. These two concepts can often coexist peacefully, with occasional conflicts but also some useful historical contingencies: during the Second World War, copper was scarce, but the government had plenty of silver, which also functions as a conductor. As one editorialist put it, "the law requires the Government to amass the silver... but does not specify where it shall be kept..." so one could "store it in electrical equipment, like that at... Niagara falls." A lot of it ended up being used in the Manhattan Project.

The book also notes that Depression-era policies of buying up silver, which was both a giveaway to silver-mining states and a way to increase the amount of liquid currency in the US, had an unintended side effect: China's monetary system was based on silver, so raising the price of silver amounted to throwing China into a deflationary depression, right around when their spending needs were rising because of an impending war with Japan. It's great to have global supply chains, but sometimes big countries export their problems to smaller ones.

Overall, it's a fun book that adds some texture to the history of the twentieth century, with weird incidents that don't make the usual history books, like the coin shortage of the early 60s, and odd characters like Henry Jarecki, a Yale-educated psychiatrist whose adventures in international arbitrage may strike some crypto-experienced readers as eerily familiar.

Born to Be Wired: Lessons from a Lifetime Transforming Television, Wiring America for the Internet, and Growing Formula One, Discovery, Sirius XM, and the Atlanta Braves: Talk of "paradigm shifts" and the like has been derided as MBA-speak for a long time, for the pretty obvious reason that whenever we're in a new paradigm there simply isn't enough information for the tedious project of finding comparable decisions and measuring their ROI, meaning that instead we're forced—forced!—to think about how if just x% of people want a given product and only y% of them buy ours, we'll have $Z any time now. But paradigm shifts are useful vocabulary, because sometimes you're scaling something that works and sometimes the people scaling a thing that seems to work are raw material for the ones who've figured out what's really going on.

John Malone handles paradigm shifts incredibly well, as is necessary in media. The general Diff theory of media ($) is that when there's a new distribution tool, the money is in figuring out how that tool works, and once the tool is figured out, the money is in having good content. So old-media brands tend to decline in the early days of a disruptive change, and rally later on. Malone started out on the new-distribution side, finding his way to the early cable industry by way of Bell Labs and McKinsey. When he joined, "cable" was just a way to deal with bandwidth limitations: there were plenty of rural communities where nobody's TV was strong enough to receive a signal, but where a bigger antenna could get a signal that could be shared more broadly. One of the first shifts he had to navigate was that cable was more than broadcast-for-people-out-of-reach-of-broadcast, but its own medium. Within that, he had to make the right bets on regulation and technology—there are parts of the book where he's making the same kinds of hardware bets that make and break careers in Silicon Valley. It's an impressive evolution: the earliest iteration of cable TV was a pretty standardized business that anyone with a bit of capital in a rural area could put together, and over time it evolved into a load-bearing part of the world's information infrastructure.

Malone isn't just famous for making money in cable and other media. He's also famous for how much of it he kept. The text of the book emphasizes one part of this that he's famous for—any time he manages to structure a transaction so it generates lots of depreciation or zero realized capital gains, he tells the reader so. The subtext is that over the course of TCI and Liberty's various acquisitions, divestitures, spinoffs, asset swaps, tracking stocks, etc., Malone tends to end up owning a bit more of whichever part of the complex is appreciating in value the fastest. This is hard to avoid given that Malone's a smart operator, and you can't fault him for making the right trades. But it does mean that shareholders who tagged along with him did better if they followed his moves exactly. (One defense here is that if he'd started a hedge fund instead of joining a cable company, and the assets in question had performed roughly the same way, he would probably have made even more from a 2&20 fee structure.)

Like Bill Gates' Source Code, Malone notes early on that he's on the autism spectrum, and that this made it harder for him to be a friendly backslapper but also made it a lot easier for him to just tell investors that he cared about EBITDA and didn't mind in the slightest if they preferred GAAP net income, and also made him ideally suited to complicated, tax-optimized financial transactions. So Malone's story is part of the story of 20th century economic growth: globalization and technological progress raised the output of people who can focus on narrow problems for long periods, relative to the output of people who were pretty good at making lots of friends. But the matching problem is harder for people on the spectrum; if you're charismatic and neurotypical, there might be a pretty small difference in outcome for selling used cars versus selling life insurance. But for someone on the spectrum, the specific complex system they end up spending their time optimizing has a big impact on how optimized it gets.

Open Thread

• Drop in any link or comments of interest to Diff readers.
• The Malone book got some help from Mark Robichaux, who wrote the previous Malone biography, but there are enough Malone-specific stories that it counts as a memoir. Which other business memoirs are most worth reading?

Diff Jobs

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

• A startup that’s beating Cochrane at systematic reviews (both in comprehensiveness and speed) and building the Github for science is hiring a product engineer who loves solving problems and wants to help index 500M articles, think through verification interfaces, and optimize the domain-specific model. A love for Factorio is a plus. (Remote, Toronto)
• A transformative company that’s bringing AI-powered, personalized education to a billion+ students is looking for elite, AI-native generalists to build and scale the operational systems that will enable 100 schools next year and a 1000 schools the year after that. If you want to design and deploy AI-first operational systems that eliminate manual effort, compress complexity, and drive scalable execution, please reach out. Experience in product, operational, or commercially-oriented roles in the software industry preferred. (Remote)
• Fast-growing, General Catalyst backed startup building the platform and primitives that power business transformation, starting with an AI-native ERP, is looking for expert generalists to identify critical directives, parachute into the part of the business that needs help and drive results with scalable processes. If you have exceptional judgement across contexts, a taste for high leverage problems and people, and the agency to drive solutions to completion, this is for you. (SF)
• Thiel fellow founder (series A) building full-stack software, hardware, and chemistry to end water scarcity, is looking for an experienced software engineer to help build the core abstractions that enable global cloud seeding operations - from mission planning to post-flight analysis. If you have 5+ years building production systems with complex integration requirements, please reach out. (Los Angeles)
• 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)
• Ex-Citadel/D.E. Shaw team building AI-native infrastructure to turn lots of insurance data—structured and unstructured—into decision-grade plumbing that helps casualty risk and insurance liabilities move is looking for a data scientist with classical and generative/agentic ML experience. You will develop, refine, and productionize the company’s core models. (NYC, Boston)

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.