Longreads + Open Thread
Reading, Crypto Cities, Tobacco, Robots, Bubbles, Bernanke
Neckar on the reading obsession, with Warren Buffett as a case study. Yes, great investors read a lot, but no, they don't get literally all of their ideas from reading. One way to read this post is that it's not about the tradeoff between doing in-depth written research and having lots of one-on-one conversations with informed people, but the tradeoff between doing both of those things, at more scale than most people can manage, without running out of energy or getting distracted by errands. The unsung hero of Buffett's career is his staff, who are apparently able to screen off enough unproductive distractions to leave time for an above-average level of 10-K perusal and an above-average level of networking.
Vitalik on crypto cities. One of the problems crypto approaches to cities will struggle with is the question of whether you want a token that has land-like characteristics, or a token-based model overlaid on cities' land-driven economics. (The formula for a prosperous city is that a dense network simultaneously raises the return for engaging in whatever activity the city specializes in, and raises the fixed cost of living there enough to price other activities out. This is a brutal but effective way to improve the quality of the network, as you can measure anecdotally through the quality of the eavesdropping.)
One of the delights of the newsletter industry is highly specialized publications covering a narrow topic with an expert perspective. The latest example I've discovered: Gene Hoots on the profitability of the tobacco industry.
Japanese construction companies have been working on "skyscraper factories" for decades, but they don't seem to have panned out. This is a great case study in the fact that "automation" is workable for a set of human activities, but doesn't pay at all for many other ones—and the reasons for this are topic-specific, not easily generalizable.
"The 2000s housing bubble was greatly exaggerated." True! It's important to continually re-revise the financial crisis to prevent something like it from happening again. The current plausible model is that there wasn't a global housing glut, but there were local housing gluts which became a global problem because the marginal additional house was disproportionately likely to be included in the same asset-backed securities as other new houses; the correlation between assets in structured products was higher than it looked, and their AAA ratings were based on the assumption of low correlations. There was a minor bubble in housing, and a bigger bubble in certainty, and when the certainty bubble popped it took housing down with it.
The Courage to Act: A Memoir of a Crisis and Its Aftermath: depending on your particular view of economic theories, it was either a very good thing or a very bad thing that the Fed chair during the financial crisis had spent some of his academic career studying the Great Depression.
As always, share any links that Diff readers would enjoy.
I’m going to be hosting a regular show on the new social audio app Callin. First episode is Wednesday at 3:30 ET/12:30 PT. This one will be an ask-me-anything episode—feel free to reply to this email with questions and I’ll answer on the show.
Productivity is notoriously hard to measure in services, but you can estimate relative differences in productivity by looking at which companies outcompete others. Are there any interesting examples of service companies that have been able to drive sustained improvement in productivity by changing worker behavior or organizational structure, rather than purely by using tools?