This is The Diff’s weekly roundup of longreads, books, and open questions. Feel free to drop in links and comments of interest to fellow readers.

You’re on the free list for the newsletter. For full access—and to read the back catalog—subscribe today!

The Diff is a reader-supported publication. To receive new posts and support my work, consider becoming a paid subscriber.

Longreads + Open Thread

Retail, Sybil Attacks, Discoveries, Roditi, Chongqing, Moderna

Byrne Hobart

Aug 6

Longreads

Spencer Soper in Bloomberg has a fun story on the Great Candy Caper: Amazon sellers who resell goods they're buying more cheaply at retailers like Sam's Club. Pricing is complicated, both for the product manufacturers and the retailers—the pricing decision that gets someone to sign up for Sam's Club, or pay a visit, is not necessarily the one that will get them to place an Amazon order, and the price hit from discounting also varies by channel. These stories tend to crop up at a particular moment in the lifecycle of small arbitrages: early enough that platforms haven't banned them, but late enough that there's a lively market in courses and tutorials on how to get an advantage.¹

Disclosure: Long AMZN.
Danny Nelson and Tracy Wang on the elaborate story of two developers who created a dozen fake identities in order to hype a financial Matryoshka Doll of DeFi entities, which ultimately controlled $7.5bn of cryptoassets (albeit with some deliberate double-counting). One of the broad problems lots of systems run into is sybil attacks, and one limitation on them in crypto is that while pseudonymous developers are accepted, it tends to take some effort to actually build and market a project. Real names are proof-of-(reputational)-stake, and pseudonyms require proof-of-work. But when a market is growing fast enough, the work required to create a fake identity isn't enough to make those fakes too costly to be worth it.²
Matt Clancy asks how common independent discoveries are, including some clever studies that try to quantify it—as it turns out, the way people do structural biology basically create natural experiments for tracking simultaneous inventions. It concludes with a somewhat anxiety-inducing observation, that the more important a given piece of research is, the higher the odds of getting scooped. Via Nintil.
Swen Lorenz of Undervalued Shares has a profile of investor Nicholas Roditi. Roditi doesn't give interviews, and has been managing his own money for a few decades, so he's hard to track. Among other things, this piece illustrates that it's possible to accumulate a staggering amount of money while keeping a minimal public profile, even if your career launchpad is one of the most famous investors in the world.
Yueran Zhang in Made in China Journal looks at the Chongqing Model in retrospect. This was a political movement led by Bo Xilai, which revived some Maoist behaviors, cracked down heavily on corruption (at least from other political factions), and then collapsed in a scandal. This was a theatrical movement—when they raised taxes on housing, tax collectors would wait at airports for homeowners who hadn't paid, with banners saying "Taxpayers, welcome back!" For a one-party state, China has an amazing level of political diversity, and this is one of the weirder experiments the country has seen recently.

Books

The Messenger: Moderna, the Vaccine, and the Business Gamble That Changed the World: In early 2020, Moderna was a company with a long track record of interesting ideas that didn't lead to any commercial products. Now it's doing over $20bn in revenue. This book is partly about the long process of exploring new discoveries and figuring out their implications, then shifts midway through to being a story of breakneck efforts to scale manufacturing and get regulatory approval. Recommended.

Open Thread

Drop in any links or thoughts of interest to Diff readers.
Moderna was a somewhat obscure company that happened to be well-prepared to tackle an emerging existential risk. Are there other companies like this for other existential risks? I'm sure people at SpaceX occasionally think about how that company might save the human race, for example. And I wonder how many companies have an "in case of runaway artificial intelligence, break glass" plan.

Reader Feedback

After last week's post on Jane Street (now the most popular Diff post of all time!), I got lots of comments, publicly and privately, on the question of whether their use of Ocaml is a brilliant move, a perverse decision, actually insignificant, etc. One of my favorites, a comment from Levi Ramsey:

A benefit of choosing a fringe language is that you can generally hire developers with less professional experience. With mainstream languages there's an impulse to hire experience so you don't have to mentor and can ramp them quickly. But if you're using a language with few experienced developers, you're going to be mentoring everyone regardless of experience, so you're going to focus on hiring for ability to be mentored.

This applies in a different direction to companies with idiosyncratic management styles, or with highly opinionated code standards and developer tooling—weird practices

A Word From Our Sponsors

Tegus is the first port of call for M&A professionals and institutional investors ramping up on an industry or company.

Get access to a database of 35,000+ expert call transcripts, spanning 5+ years, or schedule expert calls through the platform for a fraction of the usual cost.

When thousands of research analysts are pooling their expert calls into an on-demand database, using Tegus is table stakes. It's the leading platform for due diligence and primary research.

See the power of a Tegus subscription, and get up to data parity with your competitors, with a two week free trial through the Diff.

Try Tegus

1

Online courses can be quite valuable, but there's a somewhat efficient market at work: if the course teaches you a few simple tricks that instantly make money, those tricks will also be easy to detect and block. Whereas a course that teaches you something that requires effort to implement won't. A good rule of thumb is that the value of practical education is proportionate to the ratio of time required to implement divided by time spent in the course. This applies to other kinds of advice, too.

2

Sybil attacks can show up in other domains, too. In politics, astroturfing basically counts, since a large number of mildly-informed voters who suddenly care about some niche issue will make that issue look like it’s popular, even if it’s actually just being marketed well.