This is a terrific piece. You state the problems and issues clearly, and address them with much relevant data. And your conclusions are very solid. This essay should be widely read.
"Mistaking an increasingly accurate map for a theory of geography" that line is going to stick. It names something I've been struggling to articulate for a while: the difference between a system that gets better at predicting outputs and a system that actually understands what it's predicting about. The map analogy works particularly well because a map can be almost perfectly accurate and still tell you nothing about why the terrain is shaped the way it is. In my work, there's a version of this that comes up constantly. A life pattern can be mapped with considerable precision. But the moment someone asks "why does this keep happening to me," you've left the map entirely. That's not a limitation of the map. That's a question the map was never designed to answer.
Erik, nicely done. Should probably take down paywall. Corollaries to you first point are truisms in finance: profitable trades are arbitraged away. Price incorporates the new information. Arbitrage means that profitable strategies without moats have short shelf lives. Conversely, much of the point about IP is to create a legal moat, and thereby reward innovation with a bit of monopoly rent. It's a whole set of arguments about how markets work. For some reason very hard to grok, though.
This is a terrific piece. You state the problems and issues clearly, and address them with much relevant data. And your conclusions are very solid. This essay should be widely read.
Thanks. I appreciate that.
"Mistaking an increasingly accurate map for a theory of geography" that line is going to stick. It names something I've been struggling to articulate for a while: the difference between a system that gets better at predicting outputs and a system that actually understands what it's predicting about. The map analogy works particularly well because a map can be almost perfectly accurate and still tell you nothing about why the terrain is shaped the way it is. In my work, there's a version of this that comes up constantly. A life pattern can be mapped with considerable precision. But the moment someone asks "why does this keep happening to me," you've left the map entirely. That's not a limitation of the map. That's a question the map was never designed to answer.
This is the fundamental analysis from Das Kapital...
Erik, nicely done. Should probably take down paywall. Corollaries to you first point are truisms in finance: profitable trades are arbitraged away. Price incorporates the new information. Arbitrage means that profitable strategies without moats have short shelf lives. Conversely, much of the point about IP is to create a legal moat, and thereby reward innovation with a bit of monopoly rent. It's a whole set of arguments about how markets work. For some reason very hard to grok, though.