Tim Bauer

I don’t know exactly what’s going on here, but it does feel AI-related. Unlike PM and eng, which started growing in 2024 (two years post-ChatGPT), design didn’t. If I had to venture a theory, I’d say that because AI is allowing engineers to move so quickly, there’s less opportunity—and less desire—to involve the traditional design process. That said, you’d think design would become a differentiator as more products compete for attention. Something to think about for your company! We’ll keep watching this trend and AI’s impact on org design more generally. One interesting observation we made when we went a level deeper: the ratio of demand for PMs vs. designers has flipped. In mid-2023, we went from more open designer roles to more open PM roles. And ever since, PM demand has been pulling away (currently 1.27x). This will be another trend to monitor, in terms of how AI is reshaping org design.

Bill Gurley’s article doesn’t conclude “paid marketing bad”, just “be very careful”. Big outcomes have clearly come from paid marketing: Monday, Grammarly, Squarespace. He himself concludes that it “has a very important place in business… that requires.. thoroughness in its implementation” His 10 points on the pitfalls of CAC/LTV math (all true) boil down to: - Organic channels are better economics with better user: social, virality and PR - If you can’t track which customers came from LTV, you over count all new customers which you would’ve acquired anyway. Your equation is just wrong. - Current LTV math does not hold in the future as you scale. Costs scale as the business scales. You might acquire worse customers who churn quicker. Raising ARPU can cause churn. Companies use dirty tactics to prevent churn (increase LTV) and you can see ensuing complaints in the BBB - Spending it better on the customer to improve the experience and increase organic growth can be way better for long term business (eg Amazon) - You have to keep running the treadmill to grow and if your balance sheet can’t hold it, your business slows down and the valuation dips from peak (thankfully instruments like GC’s CVF addresses this and many other points here)

new Krea Edit on iOS. tap any element to magically select it and edit with the latest models like nano banana and more. try it now for free.

SBF’s League of Legends thing is actually pretty interesting. We know that Sam played a lot of the game. Here’s a screenshot from 2021 showing ~800 games in a single season (one year), which is a lot, anecdotally speaking. Each game lasts about 45 minutes, so that’s roughly 2 hours a day, every day, for a year. Sam also used to tweet frequently about the game (including about how much he played). Despite this, as you might be aware, he had an abysmally low ranking -- around mid-Bronze at the time. His username was known, so there’s not much uncertainty around this. Mid-Bronze corresponds to being lower than about 75% of all ranked players (probably charitable) -- which includes many casuals, as well as plenty of children who play the game. It used to take, if I recall correctly, roughly 10 games to get ranked in a season. Sam had nearly a thousand. So we know that Sam played a lot and seemed to care about getting better (direct quote: “I know I've said its name enough to imply I'm good at it, but I'm really not. It's actually embarrassing how little I've grown at it”.), but still fluctuated around the 25th percentile. This is fairly surprising to me, because League is decently g-loaded, and all of his other credentials imply he has meaningfully high general intelligence (Jane Street, MIT physics/math, 99.5th percentile as a reasonable base case for consensus?) If you’ve played the game, or MOBAs in general, you can probably tell that it’s decently g-loaded. The one peer-reviewed study that exists on this finds about a ~0.44 correlation between experienced League of Legends Elo and fluid intelligence (think matrix-based IQ tests). Here are some comparables so you can intuit whether that’s meaningful: > SAT ~0.82 > High school grades ~0.54 > Educational-context math achievement ~0.41 > Youth chess ~0.32 > Unranked chess ~0.32 > Ranked chess ~0.14 > Adult chess ~0.11 These numbers aren’t strictly comparable (different samples/corrections etc.), but they’re OK for intuition. So intelligence is roughly as predictive of League rank as of classroom math achievement, and quite a bit more than in any mainstream chess sample. Now, this is just one study with a sample of 56, and we’re doing serious hand-waving, so maybe that estimate is a bit off, but let’s assume for now that it’s broadly right (it doesn’t *sound* very wrong). What are the odds that someone with ~99.5th percentile fluid intelligence ends up at the 25th percentile in a game 0.44 correlated with fluid intelligence? With some naive assumptions -- roughly 2%. In that case, we should plausibly be able to find a non–cognitive-capacity source of the large residual, but I can’t really think of anything obvious. I think he played on the Japanese server, which would mean he’d have higher latency than normal, but I’ve played on comparable latencies and seen many do the same -- it usually doesn’t meaningfully move rank away from the normal counterfactual. Set aside some of the statistics for a second and move away from precision -- given all the uncertainty -- and think of a hypothetical with comparable but more intuitive tasks with similar loadings. Imagine a future MIT/JS guy who perpetually ranks at the 25th percentile in his average high school math class despite putting in effort, or one who spends hundreds of hours a year on chess and still ranks in the bottom 25 percent of a cohort (which should be less surprising given the meaningfully lower loadings). How surprised would you be? Actually, very? Maybe he had really bad motor skills? That seems unlikely for a trader, I’ve gotten tested on motor skills before in interviews. He seemed to consistently play only a few champions (characters), a lot of people underperform because they jump around too much, so that doesn’t explain it either. Some emotional/personality extremes, issues with frustration, etc.? Probably the most plausible ex post. But it would have to have been pretty bad -- the guy is bad at chess because he’s too emotional? That’s pretty bad. Something still feels off, and I assume we’re probably overweighing his g -- paired with tiger parent pedigree (2 Stanford Law profs, elite prep school) inflating the other meritocracy signals, this is probably a fair update.