[Update: Tim Worstall has a beautiful response to this post, applying the Ricardian idea to trade between humans and robots (or the human owners of robots). Will the robots out-compete us in everything, causing mass unemployment? Nope. Robots will never have a comparative advantage in everything. Once again, Tim's Ricardian point seems totally obvious once you have seen it, but was totally non-obvious (to me) until he said it.]
Noah Smith says: "Maybe people are perfectly smart and rational enough to understand the David Ricardo idea,..."
I'm not so sure.
I must have been taught the David Ricardo idea two or three times, and I must have taught it to my students about 20 times, and then this happened:
I was chair of economics at Carleton. One of the jobs of the chair is to decide (with consultation etc.) which professor teaches which course. All courses are different, and all professors are different, and teaching is important, and I thought it was really important to make sure that each course was taught by the best professor for that course.
So I went down the list of courses, assigning the professor who I thought was best for that course. (Can you see my mistake? It's not that obvious, is it?)
I got about halfway down the list of courses before I realised that something was badly wrong. I had some professors teaching more courses than they could possibly teach.
Then it dawned on me. I was allocating scarce resources (professors) according to their absolute advantage. I tore up the list, and started again, this time allocating scarce resources according to their comparative advantage.
Glenn MacDonald and Jim Markusen would beg to differ. But I think their point would be that if ECON 1000 is much more important than all the other courses (as I like to think it is), and if two identical professors teaching the same course do not teach the students twice as much as one of those professors (which seems reasonable), you might put the best prof in ECON 1000 even if that prof would be only slightly better than others in ECON 1000 but much better than others in some less important course. We need to look at relative importance of courses as well as relative abilities of professors, if things don't scale up in proportion.
(One other related anecdote from that era: for some forgotten reason I was a reviewer for the proposed Ontario high school economics curriculum. One item on the proposed curriculum was the gains from trade due to differences in absolute advantage. I said it should be comparative advantage. They said it was too hard to teach comparative advantage. I said they should teach comparative advantage even if it meant deleting everything else from the curriculum, or else not teach gains from trade at all. They thought they had an eccentric cranky professor on their hands, and would only accept my recommendation if it were supported by other chairs. Which is why, somewhere deep in the silicon records, there are emails from the chair of every single Ontario economics department saying we should teach that the gains from trade come from comparative advantage not absolute advantage. I fear that some may still be teaching the case where England has an absolute advantage in wheat and Portugal an absolute advantage in wine. And stopping there. And that's the kids who actually take economics at high school.)