- Carina Chocano profiles Duolingo founder Luis von Ahn in The New Yorker. One interesting feature of this is that von Ahn's early work focused on crowdsourcing data that machines couldn't generate on their own, like reading distorted words in scanned books, but has since moved away from that kind of work even though it's gotten much trendier. Duolingo was originally a more collaborative platform, but has become more of a one-to-many product than a peer-to-peer one—in two-sided networks there's one side that's typically harder to manage, and some companies end up solving that by building one side themselves.
- Since conservative commentators are in the news, it's interesting to revisit the career of Rush Limbaugh. These two NYT pieces, one on his rise in 1990 and one on "late period Limbaugh" in 2008. (The 2008 piece is a fascinating time capsule, with Ira Glass praising Limbaugh: "I’d notice that I disagreed with everything he was saying, yet I not only wanted to keep listening, I actually liked him.") Political commentary is a form of entertainment (even the wonks have to practice wonkery-as-entertainment, since a compelling analysis has less effect on popular views than a great slogan). One useful way to understand such personalities is that for the fans, detractors are part of the show.
(Via Philo at MD&A.)
- Morgan Stanley has a thoughtful report on the nuances of stock-based compensation. Part of what's helpful about this piece is that it clarifies that accounting adjustments that change the presentation of numbers—moving equity compensation from a footnote to the P&L, or presenting an adjusted earnings metric that excludes them—doesn't have much of an effect on investors' views of companies. In this case, it may actually be helpful that retail investors tend to be less valuation-sensitive, because it means there isn't as large a population of investors that companies can actively mislead through their disclosures.
- In Life is Computation, an argument that transformer-based models can't be intelligent because they can't indefinitely attempt to execute programs that may or may not halt. It's a surprisingly fun piece (I found that both of my objections to it ultimately got mostly satisfactory answers.) Since the specs are easier to measure for artificial intelligence than for the natural kind, it's easier to reason about their capabilities—but, frustratingly, hard to use that reasoning to accurately compare them to human intelligence since its limits and functions are so hard to figure out. AI continues to be useful both as a practical tool and as a way to clarify what the "I" means in other contexts.
- Nadia Asparouhova on models of talent scarcity. The core argument is that different assumptions about talent distribution lead to different hiring practices. And one important thing to note is that the "distribution" in question is not just a function of people but a function of the task at which they're being measured. Tournament-style tasks (sports, or getting to #1 market share in a business with network effects) will have a more skewed distribution, while tasks with more linear rewards will have something closer to a normal distribution. And this can be taken even further, since there are some companies where a single employee can make a small positive contribution to the outcome but can also make a substantial negative one in some cases. It's a good idea for companies and other organizations to think about these distributions when they think about how they hire.
- And speaking of talent, this week's Capital Gains is a writeup on which traits pay off in financial careers, and why. The industry is naturally a byword for where to work if you're purely focused on making money, but still has a term of art for people who actually try to make money. And as a reminder to new readers, Capital Gains is a spin-off newsletter from The Diff, breaking down one concept a week from the worlds of finance, economics, and corporate strategy. You can sign up here.
- Frozen Desire: Meaning of Money: An incredibly fun,beautiful, and depressing book on the meaning of money and our relationship to it. The book is partly a tour through the history of money, and our attitudes to it, as reflected in fact and in fiction (I opened the book to a random page, and it referenced Marx, Balzac, Eliot, and Luca Pacioli, the inventor of double-entry bookkeeping.)
- Chaos Kings: How Wall Street Traders Make Billions in the New Age of Crisis (coming out June 6th): Profiles of traders and academics who bought crisis insurance, with a focus on times when it paid off. The author tries to make the writing a bit closer to that of a thriller than a finance book (hard to pull off when the topic is research, and a superfluous level of excitement when the topic turns to the global financial system being on the brink of collapse). One difficulty tail risk funds have raising money is that, by design, they underperform most of the time. And one problem their investors run into is that the truly impressive numbers, like returns in 2008, single-day earnings from the 2010 "flash crash," or March 2020 performance are hard to contextualize. They're a small sample size, and they're never samples from the same distribution—every crisis has a different cause, and every crisis changes the market for crisis insurance in such a way that backtests won't be predictive.
- Drop in any links or comments of interest to Diff readers.
- Big tech companies have been mostly secular growth stories, with some small fluctuations depending on the economic cycle. But as they get more mature, and become a larger share of GDP, they'll have revenue more tied to general consumer and business spending trends. When does this happen, and what are the consequences once it does?
A Word From Our Sponsors
More and more investors use Daloopa to build and update their financial models. Daloopa captures every reported indicator—from GAAP metrics in SEC filings and press releases down to key performance indicators mentioned in management commentary, or buried in a footnote. Daloopa allows one-click model updates within minutes of new data releases. And with links back to source material, you can always double-check the math and get the numbers in context.
Daloopa is all about increasing idea velocity, and ensuring that investors spend less of their time looking for numbers and manually entering them into spreadsheets, and more time coming up with good trades.
Companies in the Diff network are actively looking for talent. A sampling of current open roles:
- A well funded seed stage startup founded by former SpaceX engineers is building software tools for hardware engineering. They're looking for their first marketing lead who will be responsible for marketing strategy, operations, and other content support. This person should be passionate about working closely with customers building satellites, rockets, and other complex machines. (Los Angeles)
- A company building ML-powered tools to accelerate developer productivity is looking for a mathematician. (Washington DC area)
- A startup building a new financial market within a multi-trillion dollar asset class is looking for a senior ML engineer, especially someone interested in using LLMs to make unstructured data more tractable. (US, remote.)
- A fintech startup that gives companies with complicated financials a single source of truth for managing their cash flows and understanding their unit economics is looking for a growth/operations associate who can effectively pitch to founders, CFOs, and financial service providers. (Bay Area, hybrid)
- A company building zero-knowledge proof-based tools to enable novel financial arrangements is looking for a senior engineer with a research bent. Ideal experience includes demonstrations of extraordinary coding and/or math ability. (NYC or San Diego preferred, remote also a possibility.)
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.