Longreads + Open Thread

Longreads

• Sophie Haigney writes in the NYT about the rise of the group chat. A fun modern media paradox suggests that social media is for middle-ground content: if it's too personal to share on social media, it goes into a group chat instead (where there isn't as much surface area for drive-by rudeness), and conversely, if it's too business-focused to share, it also goes in a group chat! (Where they can, charitably, be discussed in private, and where, less charitably, they’ll be auto-deleted before the subpoena hits) If there's ever a big data breach at Signal or Telegram (Edit: a reader notes that strictly speaking, a data breach would not have this effect. Exploiting a vulnerability in the app to capture plaintext conversations would, though), the FTC will have a wonderful time reading all those messages. (The DEA will probably get involved, too.) Another side effect of this is that it changes what's upstream and downstream in media. Even a few years ago, the sophisticated view was that TV news was important, but it was also determined by whatever was happening on Twitter the day before. But now Twitter is a step behind, and the key node for high-value virality is the group chat.
• Dan Wang wrote his latest, and last for a while, annual letter. One of the sad themes in these letters is that they're a story of China slowly closing off from the rest of the world. His 2020 letter is full of anecdotes from Beijing, but this one talks more about hanging out with Chinese expats in Thailand. His thoughts on China's government are helpful: the CCP does identify problems, and does try to address them, but it tends to solve less than it exacerbates.
• Steven Sinofsky reflects on the fortieth anniversary of the Mac. A useful exercise in tech history is to get yourself into the mindset where something like the Mac is a big deal—where the first time you use it, you can't sleep after because you're so excited about what's now possible to build. There’s one important take worth noting: "When I would bury myself in Unix systems programming it seemed more like a series of things, tricks, you needed to know. Macintosh felt like a system. As I learned more I felt like I was able to guess how new things would work. I felt like the bugs in my programs were more my bugs and not things I misunderstood." (Emphasis added.) Building a tool people can trust means building a tool they trust more than themselves. This was one of the things I noticed when we got some knockoff Legos for the kids: if you're assembling a Lego, and it doesn't work right, you misread the directions; with the knockoffs, there's a good chance they miswrote them instead.
• Dan Luu asks why good content gets posted on the wrong platform. Longform content is one of those momentum-driven phenomena where the only way to produce anything of value is to get into the right groove and stay there. So sometimes the publishing medium is less a function of where the finished product would do best, and more one of what interface is most likely to lead to a good topic sentence followed by a good paragraph followed by a clean transition, all leading to a solid conclusion. Fortunately, we're not too far from a point where "this is fine but I'd prefer it as a blog post/podcast/video" will be replaced with a browser feature that puts the content into whatever format the reader happens to prefer.
• Arpit Gupta on the hypothesis that the first cities were central nodes in the bronze trade. Bronze is made of copper (common) and tin (rarer), and it's much more valuable than either of those elements combined. So it's an early example of a situation where complex supply chains with multiple inputs were necessary. All the economic complexity and agglomeration effects we enjoy today started with some group of people agreeing to settle somewhere, and some other group agreeing to return to that place as a central point for trading one set of goods for another.
• This week's Riff covers factor-neutral investing, the tech/finance convergence, generalists, grocery stores, and more. Listen with Spotify/Apple/YouTube.
• And in Capital Gains, we talked about Hotelling's Law, a simple model that shows why competing products tend to get less differentiated over time. Like other economic models, it's too simple to describe reality, but gives us a short list of explanations for why the model doesn't match what we see.

Books

The Asylum: The Renegades Who Hijacked the World's Oil Market: The most important globally-traded physical commodity is crude oil, and crude oil prices are, of course, set by traders at a venue that was originally founded as the "Butter and Cheese Exchange of New York." This book is the story of how it happened: how that obscure exchange became the NYMEX, and how it became the main benchmark for oil. It's easy to be taken by surprise at things like the SPAC boom or IPO pump-and-dumps (note: this is still happening!). But the finance industry used to be much, much scruffier than it is today, particularly around the periphery. The story starts at a time when NYMEX didn't even trade crude futures, and its big business was potato futures. This was an incredibly corrupt business: they traded Maine potato futures, and traders were not above bribing Maine's farmers to withhold shipments or bribing inspectors to let rotting food through. There were other scams: a trader who executed a trade for a client might trade ahead of the client, or go back through their records at the end of the day and retroactively assign profitable trades to favored clients. (Connoisseurs of 90s political scandals may recall that this was the best explanation proferred for that time Hillary Clinton made ~97x her money in less than a year as a first-time cattle futures trader.) The exchange was never especially beloved, but got into bigger-than-usual trouble in 1976, when potato titan J.R. Simplot defaulted on a massive short position. The exchange was censured, and given a short period to clean up its act or shut down. Since the potato market had collapsed, they needed new products to trade, but regulators would only allow them to reintroduce products, not to add new ones.

Luckily for NYMEX, long ago they'd traded fuel oil futures, and the 70s were an exciting time to bet on energy. They reintroduced those futures, they took off quickly, and the rest is history. It's still an interesting and sometimes dubious history, of course; there are many anecdotes about events that, in today's terms, were not 100% compliant with the typical financial services company's ethics policy, HR policy, drug-use-during-work policy, alcohol-use-before-even-getting-to-work policy, etc. And there's plenty of politics. (In stories with multiple sides, the author tends to be more sympathetic to whoever talked to her the most, but this is inevitable in business books.)

One surprise in the book was how long it took for the oil industry to get used to the idea of actively trading oil. They were willing to do so over-the-counter, but the big oil companies were extremely reluctant to let anyone know that they were hedging or speculating on the NYMEX. For a long time, they operated through front companies, both to obscure the size of their trades and to avoid embarrassment, but eventually they decided to be open about it.

Overall, the book is a good reminder of two things. First, many of the professional, respectable organizations of today were much more freewheeling in the very recent past. For the veterans who were around during the wild days, it can be a bit sad to let go of some of those traditions, though all the money they made along the way certainly helps. The other lesson is that it's incredibly valuable to be the central node in a network. There was huge demand to bet on the price of oil, and NYMEX got there first.

Diff Jobs

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

• A CRM-ingesting startup is on-boarding customers to its LLM-powered sales software, and is in need of a product engineer with a track record of building on their own. (NYC)
• An alt data provider is using LLMs to scrape backends and is in need of their own backend engineer—experience building with LLMs required, web scraping experience preferred. (Remote)
• A data consultancy is looking for data scientists with prior experience at hedge funds, research firms, or banks. Alt data experience preferred. (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)
• A successful crypto prop-trading firm is looking for new quantitative developers with experience building high-performance, scalable systems in C++. (Remote)

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.

Open Thread

• Drop in any links or comments of interest to Diff readers.
• The Mac was a big deal as a widely-adopted product that changed how users thought about what computers are capable of. What are some good candidates for products that weren't widely-used, but did end up having an enormous impact?