Longreads + Open Thread
Longreads
- In 1986, Susan Lammers did a series of interviews for what would later become the book Programmers at Work. Many of these interviews are now available for free on her site, including this one featuring Bill Gates, which is most notable for how much delight Gates takes in low-level optimizations. The interview was written right at the point where the performance tradeoff from using something other than assembly language started to mean that programmers switched to C instead, but a lot of Gates' thinking is intrinsically low-level: the more you're working with individual bits in specific places in memory, the more you can enumerate every reachable state of the program and make something as optimized as it can be (an early payoff from doing Nand2Tetris is seeing just how much densely-packed cleverness is embodied by the Arithmetic Logic Unit; it turns out that there are a surprising number of things you can do with a fixed set of bits that correspond to useful mathematical operations). So in a way the interview is a tragic story, about someone who mastered a craft so well that he was in the perfect place to see that this craft was quickly reaching economic obsolescence. A Bill Gates who was solely motivated by clever code might have decided some time in the 80s to pivot into device drivers, embedded software, and other applications where counting every compute cycle paid off, but Gates was ambitious enough to move on with the times. It's as if the last verse of The Ballad of John Henry were about how he founded an extremely successful steam-powered drill company and became one of the richest people of all time. But this is the curse of all technologists: if the hardware you use keeps improving, all those tiny details you learned to notice are doomed to eventual irrelevance.
- And on the topic of inevitable doom and eventual irrelevance, take a look at Brian Potter’s recent piece on Construction Physics about what went wrong with US Steel. In the short term, taking risks is what kills institutions. In the long run, though, it's avoiding them that's the problem: US Steel was always big and important enough to justify holding off on new steelmaking technologies, and had to weigh labor efficiency at any given plant against labor troubles at all of them. It also had a dividend to think about, which makes aggressive modernization programs a tough sell. But in the end, that kind of caution still meant that the company became economically less relevant and employed fewer people—it just happened gradually enough that it was easy to ignore.
- Danny Goodwin breaks down Google VP of Search Pandu Nayak's testimony about how the search engine works. This is naturally very interesting to anyone looking to game Google, but it's also useful context for people whose view of reality is partly mediated by search engines, i.e. everyone. A lot of what makes Google special is its scale, and one reason they can be open about their use of click-based signals is that no one else has quite as many clicks. As TikTok has demonstrated, a company can get very far using mostly behavioral signals bootstrapped from showing users a partly-randomized sample of content. This is harder in traditional search, for many reasons, but it's a continuum rather than an absolutely different search process.
- Evan Zimmerman on the history of patents. It's useful context for the current state of affairs, and presents an interesting case study in the virality of standards: Venice introduced patents in the 15th century, and the first French patent was granted to a Venetian immigrant. When there are conflicting models of property rights, it tends to lead to convergence on whichever property rights model either a) creates the most wealth in the aggregate, or b) allows a few interested parties to capture more upside. This point shows up in some critiques of neoliberalism: free trade implicitly assumes an Anglo/US model of property rights and commercial law, and tends to treat deviations from that model as bugs in the system. This is, naturally, invisible to anyone who already thinks in terms of that model, and in that particular case the model does indeed work well.
- David Heinemeier Hansson on how Denmark's economy really works. To a first approximation, anyone you envy was either much luckier than you or accepted much bigger tradeoffs than you would to get whatever you envy them for. This is much more true for extreme outliers. It's a good exercise to look at what free education and healthcare look like in practice, and where they might not work. (Notably, Hansson is writing about where he grew up, but not where he lives. The marginal tax rate for a high-earning Chicagoan is 44%, not too far from Denmark's 52% maximum, but the tax rate for lower earners is lower—so the US is partly a better place to get rich because it's a better place for the not-rich to take a job that makes them richer.)
- In the latest episode of The Riff, Erik and I talked about antitrust, what's disrupting education, and in particular what's disrupting elite education. For the median student, the higher a school's academic standards are, the more important the non-academic parts—brand name and network—become. (Not true in every field, of course, and very much dependent on students' choices, but broadly accurate.) Listen on Spotify/Apple/YouTube
- In this week's Capital Gains, we revisit the idea of economic profits, with a focus on why simple microeconomic models argue them away and what that can tell you about where they come from. In the very long run, these profits do go away; at some point, social networks and GPU designers get the same return on equity as steel companies and gas stations. But it can take a long time for that to arrive.
Books
- Gold and Iron: Bismark, Bleichroder, and the Building of the German Empire: following up on the Mitt Romney biography (briefly reviewed here), why not look at another politically-influential financier whose rise required navigating religious prejudices? Gerson Bleichröder was a banker who collaborated closely with Otto von Bismarck on both matters of state finance and of personal finance. In what might come as a surprise to people unfamiliar with Congress, these tasks were closely related: Bleichröder was, on Bismarck's behalf, buying railroad shares ahead of Prussia's government buying them out, buying and selling sovereign bonds based on undisclosed diplomatic developments, and alternated between using state money to bail out private interests and doing things the other around.
Much of the story is really about the development of modern states, which, at least in Germany's case, was a matter of money—how much the government could spend, and how much control was devolved to executives like Bismarck rather than legislators. Early stories in the book, when Prussia was ambitious but not rich, involve what might be described as accounting games: finding ways to spend money unsupervised, or engineering a crisis in order to get the budget to fight a war. And the book talks about the development of a sort of Imperial German deep state: the government secretly controlled a newswire (undisclosed to the diet), which was both a source of funds and a propaganda tool.
But the real story is about what the author calls "the anguish of assimilation." (That author, incidentally, was descended from an old German family that converted from Judaism to Christianity a few generations before) In Asian Godfathers, Joe Studwell notes that one reason so many South Asian countries had a Chinese business elite was that politicians didn't see them as a political threat, but did see close relationships with business as useful; the CCP's attitude towards some tech companies in China indicates that this remains a live concern. Similarly, Bleichröder was powerful and influential, but was always limited because of his background. He could be indispensable to the Iron Chancellor, but would never be chancellor himself. This made Bleichröder’s position safer than that of other Bismarck associates, but also led to constant low-level passive aggression from the Prussian establishment—jokes, spurned invitations, boilerplate regal proclamations carefully rephrased (the king's letter ennobling Bleichröder dropped a standard reference to the recipient being from a "good family").
The book is a snapshot of a particular time in history, but it emphasizes some things that never change: foreign policy is partly tied to what's going on domestically, governments try to influence the media to spin their story, "government" is not a monolith but a collection of people whose interests are often at odds, etc. Bismarck even talks about reflexivity, noting that war rumors affect market prices and that market prices, in turn, affect the likelihood of war ($, Diff).
Open Thread
- Drop in any links or comments of interest to Diff readers.
- What are some other good examples of people who mastered a craft and then profited from the transition to mass production, the way Gates did in software? Henry Ford naturally comes to mind, as well as the people who used to make trades and now run "alpha factories" instead. Who else?
Diff Jobs
Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:
- An early-stage startup aiming to reduce labor costs by over 80% in a $100bn+ industry is looking for a digital fabrication lead with a background in 3D fabrication and or architecture. (NYC)
- A well funded seed stage startup founded by former SpaceX engineers is looking for full stack engineers previously employed by Anduril or Palantir. (LA)
- A well-funded startup is building a platform to identify compliance risks associated with both human- and AI-generated outputs. They are looking for a frontend engineer with React/Typescript experience to join their team of world-class researchers. (NYC)
- A concentrated crossover fund is looking for an experienced full stack software engineer to help develop and maintain internal applications to improve investment decision-making and external applications to enable portfolio companies. (SF)
- The leading provider of advanced options analytics — “the ASML of options trading” — is growing rapidly, very profitable, and looking for a generalist who can excel in chief of staff and business development functions. A trading, quant, or similarly technical background is a big plus. (Connecticut, NYC)
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.