A friend from Argentina once told me Argentina keeps its best wines for themselves and exports the mediocre stuff, even at the sake of profits.
Similarly, a friend from Turkey once said he couldn’t find good Turkish olives outside of Turkey because “Turks are terrible businessmen and keep the best olives to themselves.”
These are anecdotal and might be untrue but I liked the idea.
At an individual level, it’s irrational to cooperate in a prisoner’s dilemma yet experiments show people cooperate.
Contributing to open source projects may fall into this category.
Have you observed any obvious behavior that goes counter to profit maximization? Any cool examples?
My father once told me of an old IBM machine, I think it was the System 3 model 15D or one of its contemporaries, or maybe the original System 38. It had some amount of memory, like 32k of memory (I’m going to get these numbers wrong), and to upgrade it you could spend many thousands of dollars to have IBM come install a control board to upgrade it to 64k. The memory was already physically in the box; they manufactured and delivered it to the customer, and sold the memory control board as an exorbitant cost option, when it was the RAM (it might have even been core storage) that was the expensive part to make.
To a lesser degree, I’ve been hearing about cars that install cost options on all models, but they don’t hook them up on the lower tiers. Like apparently all Lotus Exiges have power mirrors, they’ve all got motors in them, but they don’t give you the switch unless you pay for it. You can go to a Ford dealership, buy the right switch and just pop it in and it’ll work. I suppose it can make some sense to reduce part counts, but it’s getting to the point where it’s "we installed the option in the car, it’s hooked up, it’s perfectly functional, we’ve already put in the expense, and we’ll allow the software to turn it on if you pay for it.