Longreads

• Some interesting looks at protein diplomacy and wheat diplomacy ($). The US has been in a very fortunate position historically in that we have local access to most of the natural resources we need, and, during the post-70s/pre-fracking interregnum, we still had the world's best navy. So the US can think of resources in terms of dollar constraints, rather than the inability to acquire something at any price. (Recent supply chain issues aside—and while those have lasted a surprisingly long time, it's hard to make a case that the supply of chips for the auto industry, or other inputs, will be low indefinitely).

• McKinsey highlights the first org chart, interesting both as a historical document that shows the effects of technology on institutions, and as a very well-done infographic.

• Comedy is Preparation for Death: a very meta meditation on Norm Macdonald, laughter, and death.

• YouTube's content recommendation, explained. One thing that stands out is that as content recommendation systems mature, they get more judgment-intensive faster than they get more data-intensive.

Books

• Discourses on Livy: The Prince is a famously cynical book about how to grab and hold on to power, with lots of frankly sociopathic advice about whom to murder and why. It has its appeal for some people. The Discourses is not quite the opposite of that, but it is a gentler guide to the question of how to run a stable and non-tyrannical state. Being "Machiavellian" is probably correctly-rated: most people think it's bad, but also think it can work, at least for a short time. But the standard-issue Machiavellianism is about 99% The Prince and 1% everything else. So the question is: was he being more brutally honest in the more brutal book, or was it more of a feint?

• Thoughts on Machiavelli by Leo Strauss aims to answer this, and does so at great length. Strauss is brilliant but a little too willing to entertain crazy theses, so this book goes back and forth. There are clever insights: Machiavelli dedicated The Prince to an existing ruler, but dedicated Discourses to some friends who he thought out to be rulers—and he notes that you never can tell whether a bad person will end up running things because of random luck. Hm! And there are also more numerological ruminations on number chapters, figuring out why some quotes are in Italian and others in Latin, carefully parsing omitted details, etc. The extreme Straussian reading is a bet that a given book has a great deal of deep thought that's been carefully encoded in it, and also a bet that the author never got sloppy and quoted something from memory without double-checking. And reading Strauss is an exercise in figuring out which details Strauss discovers are genuine insights, and which are the product of being too willing to see patterns when there's just noise.

Open Thread

• Drop in any links or books of interest to Diff readers.

• I'm planning on writing a bit about the CTV market soon. Readers with strong views on where that's going and how the economics will shake out are encouraged to ping me.

• There are some very interesting experience curves out there for products like semiconductors (the most famous), solar panels, genome sequencing, and batteries. And there are some products that got cheaper for a while and then stopped; the cost to get a kilogram into orbit dropped fast (albeit from infinity), stagnated for a few decades, and then started dropping again, for example. Are there any other good examples of experience curves that have faded out recently?

Leave a comment