notes-econ-marketAsCognition

private information, Kantian free will, and the Rubino theory of consciousness

Kant had a lot to say about free will but my interpretation of some of it is that, sure, from the point of view of physics, behavior may be deteministic (or deterministic + some simple quantum randomness), but from the abstract point of view of people as entities, they can be thought of as having free will, and that's no contradiction.

A cognitive system might have various models of itself and its plans might be affected by these models, eg it might be aware that it is thinking of doing something, that it has decided to do something, etc. The Rubino theory of consciousness is unfinished but my interpretation of some of it is that a cognitive system may be so complex (and possibly chaotic) such that even if it has an intention to do (or refrain from doing) something, and even if you have a complete copy of its state at a point in time after which it has formed this intention, in general there is no easier way of predicting its intentions than simulating it and seeing what it does (or simulate it being asked its intention and answering the question, assuming its answer is honest) (note: i talked to Mr. Rubino about this and this idea was not an accurate representation of this theory; his idea was not that it must be categorically easier to simulate it than to make a simpler model, even if you have complete state and infinite computational power, but just that, given the partial information about system state and computational power you have at your disposal, you can't make a complete model). In this case, the cognitive system may be said to have 'private information' about its own intentions (note: i talked to Mr. Rubino and he thought the phrase 'private information' is misleading; he would say, rather, that what i'm talking about is 'abstracted information', namely, information that you don't have in your model so, from the perspective of your model, you attribute to the 'volition' of the entity). Consciousness may be defined as the attribute of possessing such private information (or perhaps the capability of having such private information).

Consider phenomena in economics where the principal causal factor in some market movement event is said to be the people's expectations of future market movements. For example, https://en.m.wikipedia.org/wiki/Causes_of_the_Great_Depression#Expectations_hypothesis claims "after years of deflation and a very severe recession, important economic indicators turned positive in March 1933, just as Franklin D. Roosevelt took office. Consumer prices turned from deflation to a mild inflation, industrial production bottomed out in March 1933, investment doubled in 1933 with a turnaround in March 1933. There were no monetary forces to explain that turnaround. Money supply was still falling and short term interest rates remained close to zero. Before March 1933, people expected a further deflation and recession so that even interest rates at zero did not stimulate investment. But when Roosevelt announced major regime changes people began to expect inflation and an economic expansion. With those expectations, interest rates at zero began to stimulate investment as planned. Roosevelt's fiscal and monetary policy regime change helped to make his policy objectives credible. The expectation of higher future income and higher future inflation stimulated demand and investments. The analysis suggests that the elimination of the policy dogmas of the gold standard, a balanced budget in times of crises and small government led endogenously to a large shift in expectation that accounts for about 70–80 percent of the recovery of output and prices from 1933 to 1937. If the regime change had not happened and the Hoover policy had continued, the economy would have continued its free fall in 1933, and output would have been 30 percent lower in 1937 than in 1933."

If this sort of thing occurs, then there is a puzzle. https://en.m.wikipedia.org/wiki/Causes_of_the_Great_Depression#Expectations_hypothesis says that an external event, Roosevelt's policy changes, caused people to expect inflation and economic expansion, and that these expectations, not the policy changes themselves, drove expansion. In other words, the postulated causal chain is (External event --> expectations of market rise --> market rise). This would make sense if people's internal model of the market were such that they were expecting the policy changes to directly cause the economy to get better, ie if people's internal model were (External event --> market rise). But if it is really the case that expectations were the only critical factor, not the policy changes themselves, then one would expect that eventually, after enough economic history transpires, people will learn this and adjust their internal cognitive models to (External event --> expectations of market impact --> market impact). But once people start thinking like that, it is unclear how to make the determination of which external events affect the market, because one must look not at their effect on actual economics, but rather upon their effect on people's expectations; but people's expectations are based on the result of the same cognitive process which is being undertaken here, so this is like a recursive subroutine call, like induction with no base case. Unlike a computer program in which this would simply cause an infinite loop, and the computer would hang, here the market will either move in response to the external event or not; nontermination is not an option. So this is similar to the Rubino 'private information'; so under the Rubino theory, we might consider Mr. Market to be conscious. (note: I talked to Mr. Rubino, and he pointed out that in reality, it's likely that when we say 'the market reacted in this way because people expected the market to react in this way', what probably happened was that some subset of people had other, more solid, reasons for believing that the market would react in this way, and information about their expectations influenced the market in both endogenous and exogenous ways, but who those people were, how many of them there were, and what those reasons were, we don't know, and we also don't know what their other reasons were, so we encapsulate our ignorance by ascribing the action to Mr. Market)