Byrne Hobart
Longreads
- In Foreign Policy, Tanner Greer and Nancy Yu write about China's plan to pull ahead of the US scientifically. All countries need some source of legitimacy; a recent coup, revolution, or regular elections will do the trick, but so will significant economic growth. China has been relying on that growth for a long time, but that’s harder to achieve today since they’re much closer to the frontier in many technologies. But since the country modernized so recently, they still have plenty of places to deploy what they discover.
- John Nye on technology increasing perceived inequality. The title of the post made me think I'd disagree with it, but "the problem of inequality" being referenced is not just about the distribution of resources, but about the perception of how equal and just that distribution is. If you're modeling the impact of technological progress on actual inequality, you care about three categories: inequality goes up for frontier technologies because they go from unavailable-to-anyone to only-for-the-rich; inequality goes down fast for the previous generation of frontier technologies (a mobile phone used to be a status symbol!); for older products, the effect is ambiguous. Some advances increase equality just about everywhere: food and clothes are astonishingly cheap by historical standards, but, as the essay points out, housing tends to get more expensive since the value of colocation goes up for the highest earners. (Via Marginal Revolution.)
- In the NYT, Michael Steinberger asks if the 401(k) was a mistake. There are some decent points in this piece: it's a part of the tax code that redistributes wealth to the richest, but doesn't have the benefit of moving capital to the most efficient capital allocators because 401(k)s tend to get invested in large funds that don't outperform. But what it misses is that 401(k)s also redistribute current consumption in the other direction, since the people putting the theoretical maximum of $76,500 per year into those plans are not spending the money on goods and services that other people would choose to buy. Meanwhile, people with lower savings are, on average, going to be people who prefer present consumption to future earnings. (Yes, there are other causes, like not earning much money—but unless there is some positive correlation between propensity to spend a marginal dollar and ability to earn more money, there will tend to be a negative correlation between personal savings rate and personal discount rate.) In that sense, the real problem is that opt-in retirement plans enable people to better express their preferences for consumption or savings—and a preference to spend money now and figure out retirement later is just not a good one to have.
- Sam Hammond presents 95 theses on AI. I do not endorse all of these, but he's taking a measured approach. For example, he argues that eventually open-source models will be dangerous, but that the open-sourcing of current models has been a very good thing.
- Maggie Harrison Dupré writes in Futurism about AdVon, a company that places AI-generated writing with AI-generated author photos and bios on reputable sites. There's a pattern in declining industries where companies try to monetize their brand at the cost of rapidly degrading it: in the short term, putting the Sports Illustrated logo on this content makes the content more valuable, but in the long run it makes SI worthless. That happens faster when the content provider is misleading companies about how it writes its content—the article has some fun examples of reviews for lifting belts suddenly veering into talking about Gucci, and reviews for microwave ovens that accidentally describe real ovens by reassuring customers that they can indeed use them to heat items wrapped in foil. And there turns out to be another layer to the grift, where the same company that's getting paid to produce AI-generated reviews is also getting paid to insert particular companies' products into those reviews.
- In this week's Capital Gains, we look at the quest to find a dumber counterparty. Business and finance are relative games, and while you can improve your skill at both, it's often easier to find a place where your competitors are less skilled. (Or where they're operating under constraints that you aren't—there is still money to be made competing against people who are smart, resourceful, and so levered they have to obsess over weeks and quarters rather than decades.)
- And on this week's episode of The Riff: applying Costco's economics in unusual places, Martingale strategies, and GDP-linked bonds. Listen with Twitter/Substack/Spotify/Apple/YouTube.
Books
Understanding Media: The Extensions of Man: There's always a viable niche in the media ecosystem for The Book That Explains it All. In narrative monoculture, this is often a religious text, but there's also room for sweeping nonfiction: The Selfish Gene, Thinking, Fast and Slow, Sapiens, etc. Understanding Media is an early entry into this genre. The book was apparently originally intended for high school students, and I'm sure a few high school students get a kick out of it, but it's much less a practical overview of the media environment as experienced by the average person and more an attempt to see how every medium influences the zeitgeist.
It's a zeitgeisty work! When I read The Money Game, I wondered where the book got its peppy, casually-overintellectualizing tone, and it feels like Understanding Media was patient zero for that. (Both books even compare someone to Captain Queeg from The Caine Mutiny—when a sufficiently influential book makes some pop culture references, those references will at least temporarily enter the general canon.)
McLuhan's thesis is captured in his catchphrase that "the medium is the message." This is a pretty maximalist take: I tend to think that CNBC, Stocktwits, and the Wall Street Journal have a lot in common with one another, and are different from the Hallmark Channel, Weird Sun Twitter, and Proceedings of the National Academy of Sciences. But he's right that the mode of consumption matters, and he's very right indeed when he extends the concept of "media" to include houses, clothes, money (it's the high-order bit ($)) etc. It's a very useful exercise to think about how the same message has a different impact in different media: a retweet is faster than a review, it's easier to remix a song than a novel, a folk tale never stops evolving while a slogan stays fixed but can change its meaning. (The phrase "well-behaved women seldom make history," for example, was coined with roughly the opposite meaning of what it means on a bumper sticker.) But the message still matters! The same medium can pacify or inflame people depending on its content; TV in McLuhan's age could provide endless light entertainment that distracted people from social issues, or it could bring Vietnam and Selma to every living room in the country.
Some parts of the book are quite prescient; he was early to what's now a management consulting cliche, that companies are in the business of solving customers' problems, not selling specific products. He has a riff on how every single purchase a consumer makes is fed into elaborate systems that predict demand and alter supply, which wasn't necessarily true back then but is mostly the case now. On the other hand, he thought that the EU would cause World War Three, and that TV ads would eliminate sales as a profession. Some points the book makes are weird, and lead to less wondering about whether or not McLuhan was on to something and more wondering just how dilated his pupils were while he was writing this book. (It feels more like an amphetamines-influenced work than anything else, but McLuhan was good friends with Timothy Leary.)
The trouble with reading any influential book a few generations too late is that you've already heard all the big ideas. And when you heard them, they sounded like clichés, because you've taken for granted the conclusions that are downstream from those original ideas. Reading a book like this is a good calibration exercise—people with big, influential ideas often get a lot wrong along the way. But you can invert that: as long as you get a few big things right, people will mostly forget about the rest.
Open Thread
- Drop in any links or comments of interest to Diff readers.
- Are there any forgotten works that tried to be the Understanding Media or Sapiens of their age. Or, even better, books that were recognized this way for a while and then faded?
Diff Jobs
Companies in the Diff network are actively looking for talent. See a sampling of current open roles below:
- A startup building the world’s most performant parallel-EVM is looking for a low level engineer with C++/Rust/CUDA experience. (Remote, EU preferred)
- A blockchain-focused research and consulting firm is looking for an infrastructure engineer to secure their clients’ networks. Deep experience in DevOps, Linux systems, and Infrastructure-as-Code required; previous crypto experience preferred. (Remote)
- 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)
- A company building the new pension of the 21st century and building universal basic capital is looking for a frontend engineer. (NYC)
- A private credit fund denominated in Bitcoin needs a credit analyst that can negotiate derivatives pricing. Experience with low-risk crypto lending preferred (i.e. to large miners, prop-trading firms in safe jurisdictions). (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.