On Wealth and Money

My wife shared this nice piece from Brainpickings with me, quoting Alan Watts on wealth and money. I think his "man on a desert island" definition of wealth is striking, and reminds me that wealth is context dependent, like Shannon's "information" (see also).While information theory is brilliant (I've used it alot in my work), I hate the name, in that it has separated information from meaning. It does not matter what messages "mean" in the Shannon framework, semantics are ignored.Even Shannon's analogy of information content of a message as it's level of suprising-ness is troubling, in that a high-information-content message from a known space of possible messages is only surprising in a combinatoric sense: it's only related to truth uncertainty, not semantic or ontological uncertainty, in the sense of Lane and Maxfield.That's not to say information theory is without value: as a basis for measures of complexity, it might help indicate when a system is at the edge-of-chaos, thus providing some indication of the level of ontological uncertainties one might expect from systems. In this sense, it's much better than most uses of dubious probability theory in AI.Sorry, that's a long way around to saying something about wealth and money. My friend Ben Cole has often said that money is a form of information, but what Watts is saying is that money, and in fact all forms of barter or currency are context dependent, and thus not context-independent measures of true wealth. So, money is information in the sense of Shannon information: it removes the context of meaning that defines wealth.This also relates to the basis of taxation: we should be striving for progressive taxation on wealth, which is defined by the context of our society, and requires careful human crafting around that principle. By focusing on flows of money (for instance, income tax), we miss the point: ask yourself how much real wealth you have, in the context-dependent sense of Watts, when compared to those richer and poorer than you. Taxation rates should progress on that basis. With that principle in hand, we could begin to much more meaningfully address the horrible realities of mounting economic inequality. Pretending that it's all a semantics-free calculation is the core problem, as it is in economics and governance across the board.