Longreads + Open Thread

Longreads

• Rohit Krishnan on "Seeing Like a Network," and how the world changes when people are more densely connected. That's not just a handwavy description—there is quantifiable evidence that the network of email senders in 2007-8 had more average hops between random users than the network of social media users in 2016 (6.6 vs 3.6). Dense networks mean that ideas spread faster, but they also mean that ideas are more selected for how easily they spread. So if you've wondered why the world seems to be falling apart, but your peers are also so much more successful, and your political opponents have gotten so much dumber—it's all because you, your friends, and your friends' friends have too many friends and are too good at communicating with them.
• Nima Badizadegan writes about five-nines problems—software products that need to be ridiculously reliable, fast, or correct. This is difficult for the same reason that any other effort to eliminate edge cases is difficult: the easy optimizations are very easy, and the weird cases are hard to brute-force. (This leads to its own five-nines problem, of debugging debugging.)
• In Stereogum, Zach Schonfeld asks why so many arena tours are failing to sell, or getting canceled entirely. One of the interesting details there is that promoters have more data than ever before, through Spotify streaming counts and the like, but it's not easy to measure the conversion rate between zero-marginal-cost streams and $100-$1,000+ tickets. There's also the general problem of cultural fragmentation: better recommendation algorithms make it easier to find exactly what you like, which makes it less likely that you'll like the same kind of music other people in your city do. The equilibrium might be that there are longer tours with more stops at smaller venues. It would be unfortunate if streaming started out by reducing artists' earnings from recorded music, but also ended up reducing the upside from playing live, too.
• Dan Williams on disinformation, and why it's not an especially big deal. The areas where people worry the most about disinformation are the contexts where there's a politically salient issue—i.e. a case where there's widespread disagreement about what model best explains the world and which facts ought to be salient—and this means there's a large set of news stories that will look like disinformation to someone who thinks they're highlighting rare events to make them look common, but will look like plain old information to anyone who thinks those events are too frequent. The demographic that likes actual fringe content is small, and already had weird beliefs before it got easier to manufacture and distribute them. And there's just a lot of subjectivity in deciding what counts as a defensible narrative and what's a straightforward lie. (I recently sent a salty email to the author of this Politico piece, because the piece refers to an experiment failing, but later clarifies that the experiment didn't actually happen at all. But don’t worry! The author assured me that the "experiment" that failed was one whose hypothesis was "We will get permission to do our science experiment." Which, of course, is just not what any reader would think it meant. But there is a technically not-lying interpretation of the article, so they will not be issuing a correction. Politico does, incidentally, offer plenty of coverage on the broader misinformation issue. Fun beat.)
• Brad Setser offers a grim assessment of the state of deglobalization. As measured by bilateral trade between the US and China, it's receding, but this is partly because China is shifting low value-added manufacturing to other countries while manufacturing more high value-added components domestically. That's a bigger threat, and harder to decouple from.
• This week on The Riff, we talked about meme stocks, conferences as a microcosm of city economics, and whether AI is bad for therapists or very, very good. Listen on Spotify/Apple/YouTube.
• As a periodic reminder, The Diff has an associated AngelList syndicate. If you're an accredited investor interested in early-stage deals, you can sign up here to see them.

Open Thread

• Drop in any links or comments of interest to Diff readers.
• The Setser piece is a good entry in the genre of "problems that got solved mostly in an accounting sense, but are worse in a real-world sense." Are there any other good examples of this?

Diff Jobs

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

• A mission-driven ed-tech startup is deploying AI tutors to revolutionize education. They have a strong customer base in Latin America and they're in need of a driven founding engineer. Key skills: TypeScript, React, Node.js, Postgres. (Remote, asynchronous)
• Well-funded deep tech startup that's solving water scarcity is looking or a software engineer. Experience with Deck.GL and Typescript is ideal, Python also helps. (Los Angeles)
• A top prop trading firm is looking for an intellectually curious, mathy generalist to work on projects spanning business strategy, technology, and markets. (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 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)

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.