Longreads + Open Thread

Longreads

• Patrick McKenzie on the economics of deposit insurance. A good example of the general rule that any otherwise-inexplicable behavior by big companies has some kind of explanation, either a weird incentive structure for individuals that leads to collective irrationality, or an unusual set of tradeoffs that makes an irrational-seeming behavior sensible in context. The existence of large uninsured deposits is a test case for this, and one that turns out to be illuminating.
• Chamath Palihapitiya has an annual letter covering 2022. The detail that got a lot of attention from this one was his remark about leverage ("A small line of credit that we used to drive incremental returns during ZIRP quickly ballooned as some of the assets we used to collateralize it saw up to a 70% reduction in price. What initially seemed like access to free money became a liability that we managed carefully so we could continue to do business as usual... I had to recalibrate my original models and question from first principles the purpose of leverage in the first place.") But another running theme is that some of the 2010 themes are still going strong: the letter argues that the marginal cost of energy and compute are going to zero. They can both decline, certainly—but only with the kind of capital expenditures that get underwritten much more carefully in a nonzero-interest-rate world.
• In Inverse, S.I. Rosenbaum tells the story of the life and death of TropeTrainer, a software product used to teach people to chant from the Torah. Building the software was very much a full-stack process, involving using an old text-to-speech product called DECtalk, and then modifying it to handle the correct phonemes and pitches, and adjusting it for regional differences. Software can be understood as just code, but it's really part of a more complex interaction between programmers, users, and the external world that the software references, models, and affects. And when the creator of a product dies, that entire system frays.
• Tomas Pueyo models when AI will take your job. The basic model is that an increase in productivity raises output and lowers prices, and in some cases those lower prices increase demand enough to raise employment, at least until the market is saturated. So one way to look at this is to focus on industries where demand appears pretty much infinite, like healthcare and entertainment. The piece argues that outcomes in law will get a more skewed distribution, with a handful of superstars and a large number of people who can't compete, but this tweet from a lawyer makes a compelling argument against that.
• Kit Chellel in Bloomberg profiles a gambler who developed an edge in roulette. Stories about advantage gamblers are always inspiring. The gambling industry is literally in the business of designing games that are random enough to be fun, but where the house has a persistent edge. The fact that some of these games get solved, for various definitions and over variable time periods, is good evidence that markets will never be perfectly efficient.
• This week's Capital Gains covers "Schmuck Insurance," the practice of structuring the sale of an asset in order to hold on to some of the upside.

Books

• Trillions: How a Band of Wall Street Renegades Invented the Index Fund and Changed Finance Forever: A surprisingly engaging story, given that the fundamental plot is that a group of investors and academics aggressively gave up on selecting individual stocks in order to focus on cheaply replicating overall market returns instead. As it turns out, this required a lot of work and a fair amount of interpersonal drama. One thing the book makes clear is that there was a chicken-and-egg problem that led the indexing business to have many false starts and some slow initial growth: creating a portfolio that replicated the S&P 500 required a much larger initial investment than, say, creating a portfolio of fifty stocks selected by an individual manager. So the minimum size for an index fund was larger than the minimum size for an active one. (Of course, today it would be trivial to come up with a less diversified portfolio that mostly replicated the index's performance with little tracking error, but at the time programmers were scarce and computing was expensive.)

Open Thread

• Drop in any links or comments of interest to Diff readers.
• A fair number of research tasks can be summarized as "read the equivalent of fifty books in order to find a dozen important sentences." Is anyone doing anything interesting a) automating the versions of these tasks that were always worth doing, or b) identifying new ones that are worth doing once the cost is sufficiently low? There's a set of knowledge-work tasks that are suddenly worth doing purely to provide training data, and now that tools for converting that training data into a product are more widely-distributed, it's going to cause some interesting changes.

A Word From Our Sponsors

Wealthfront pioneered the robo-advisor industry and quickly became a leading investment platform for retail investors.

Now they’re expanding what the platform can achieve. With their new Stock Investing Account, you can purchase fractional shares with zero commissions at a $1 minimum, and their new stock collections feature can help you discover, evaluate, and invest in companies you may not have already considered (such as “Apple Suppliers”, “Cloud Computing”, and “CRISPR Technology”).

To get started, visit wealthfront.com/thediff.

Diff Jobs

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

• A VC backed company reimagining retirement wealth and building a 401k alternative is looking for product/GTM/bizops generalists. (NYC)
• A profitable startup is looking for sales reps to market its AI-based services that help small companies accelerate their growth—especially people who are excited to use AI tools to accelerate some of their own work. (SF)
• A successful crypto prop-trading firm is looking for new quantitative developers with experience building high-performance, scalable systems in C++. (Remote)
• A company building ML-powered tools to accelerate developer productivity is looking for a mathematician. (Washington DC area)
• A fintech company using AI to craft new investment strategies seeks a Brokerage Engineer / Trader with 2+ years of experience in trading or operations for equities or crypto. This is a technical role—FIX proficiency required, Python skills a plus. (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.