Longreads + Open Thread

Note: We've revamped the application form for Diff Jobs. If you're thinking about trying new roles, please take a look!

Longreads

• Elad Gil asks where value accrues in AI. One valuable piece of this post is the introductory look at how value was captured in previous technology shifts: the Internet created lots of big companies, but the shift to mobile made many big companies bigger. (The exact breakdown is debatable, but Meta and Google alone make mobile a shift that has created hundreds of billions of dollars for companies born for desktop browsers.) It's a good exercise to ask whether a new technology is more valuable as a standalone product or as a complement to an existing one. That doesn't have a big effect on how transformative it is, but does determine who benefits.
• Ryan Khurana speculates on the impact of AI on advertising. An important point here: generative ads can use the surrounding content as context, which somewhat offsets the recent loss of precise targeting data. So, in answering Elad's question above, one place where AI startups might capture value relative to incumbents would be in rethinking advertising as a process of opportunistically creating new ads rather than opportunistically targeting existing ones.
• Huawei founder Ren Zhengfe in the early 1990s, and he wrote a travelog for the company's employees. Zichen Wang at Pekingology has a translation. It's interesting both as an outsider's view of America (Dallas, for example, is "similar to Inner Mongolia," and Las Vegas is "perhaps the most beautiful city in the United States.") It's a very favorable view of America, and that's a useful lesson: countries sometimes face the fiercest competition from new competitors who are trying to emulate them.
• Ben Thompson at Stratechery has an overview of the current situation with China and chips, with a focus on relative positions in different categories of chips. Most of the focus in the chip debate is about leading-edge chips made by TSMC, and the possibility that such chips might be a) made by somebody else, b) made somewhere else, perhaps by TSMC, or c) not made at all, if an earthquake, embargo, or invasion disrupts Taiwan.
• Sebastian Rotella and Kirsten Berg at ProPublica profile a money launderer who solved a two-sided network effect problem by identifying people who had US dollars in cash (i.e. cartels) and people who wanted them (people trying to convert wealth in China into more mobile forms of wealth outside of China). This story can be read as the evil twin of many other stories on cracking a two-sided network problem by figuring out which side is the harder one to acquire, and making that side's economics as favorable as possible. The article quotes a DEA agent: "With the Colombians, it had been an 18% to 13% commission [for money laundering]... The Chinese are doing it for 1 to 2% on average."

Books

• The Pyramid of Lies: Lex Greensill and the Billion-Dollar Scandal: Business books about scandals often have a slight gonzo journalist quality. Roughly midway through the book, the narrator descends from omniscience to become a character who interviews the protagonist, sometimes gets threatened by lawsuits, and sometimes gets followed around by private investigators. This should color their perception, but sometimes the subject is either charming enough or interesting enough that a side character in the narrative can still tell a good story. Pyramid of Lies recounts the story of Greensill Capital, which was pitched as a revolutionary financial services company that enabled trade finance at scale, but wound up being a lender that made bad loans and misrepresented them. It raises some good questions about what genuine financial innovation looks like: in some areas, it's much harder to do novel things inside legacy companies, but in a business like lending, big changes are often the cumulative result of lots of incremental adjustments. So a decent share of the real innovations in lending will happen within the incumbents, because they have more to lose.

Open Thread

• Drop in any links or comments of interest to Diff readers.
• Among companies that have returned to the office, what are the most useful things they've learned from companies that went hybrid or remote? And among full-remote companies, what are the benefits of office work (if any!) that took a couple years to become apparent?

Reader Feedback

A comment from Antoinette Uiterdijk on last week's link to a story on landlords using rent optimization software (which optimized rents way up):

Smaller landlords are often not just eager for the highest rent, but are equally or even more interested in getting good tenants. People who are responsible, who pay every month on the dot, who stay in the dwelling for a longer period, who are willing to do small maintenance themselves, even add some improvements, who keep the yard nice, do. to cause problems with the neighbors, etc. Renters who tick those boxes are probably worth some "discount".

This is something to think about broadly: if every transaction has some easily-measured justifications and some more subjective ones, then any systematic measurement is naturally going to find a bunch of outliers that have a strong but hard-to-measure qualitative reason for existing. (One big outlier in the Austin market, for example, is that my kids pay rent 100% below market rate, at least until they can afford more once they start their own Substacks.) Optimization, whether it takes the form of PE rollups or pricing software, will tend to miss these, and can sometimes make the market less efficient in hard-to-measure ways.

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.

Diff Jobs

• A new service that's trolling the dating market with a better product and better monetization is looking for a full-stack founding engineer. (Los Angeles)
• A high-growth startup building ML-driven developer tools is looking for someone who enjoys researching the theoretical capabilities of software systems; a good role for someone in academia who'd like to try something else. (Northern Virginia)
• A well funded early stage startup founded by two SpaceX engineers is building the software stack for hardware companies. They're looking for a backend engineer who has experience building low latency products that can ingest lots of data.
• A firm using machine learning to customize investments is looking for a data engineer. (NYC)
• A company that's building a new hyperlocal bricks-and-clicks business seeks a market launcher who can close deals with small businesses and handle rapid scaling. (If you've ever wished you could have launched markets for Airbnb or Uber/Lyft in the early days, this is your shot...)

If you’re 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.