setembro 22, 2014

Orthogonal quotes

"What happens, happens," Carla offered gnomically. "Everything in the Cosmos has to be consistent. All we get to do is talk about it in a way that makes sense to us"


[...] imagine the time when wave mechanics powers every machine and everyone takes it for granted. Do you really want them thinking that it fell from the sky, fully formed, when the truth is that they owe their good fortune to the most powerful engine of change in the history: people arguing about science.


The cosmos is what it is. The laws of optics and mechanics and gravity are simple and elegant and universal... but a detailed description of all the things on which those laws play out seems to be nothing but a set of brute facts that need to be discovered individually. I mean, a 'typical' cosmos, in statistical terms, would be a gas in thermal equilibrium filling the void, with no solid objects at all. There certainly wouldn't be steep entropy gradients. We've only be treating the existence of such gradient as a 'law' because it was the most prominent fact in our lives: time came with a arrow distinguishing the past from the future.

Greg Egan, Orthogonal (books II & III)

setembro 01, 2014

Primitive man, aware of his helplessness against the forces of Nature but totally ignorant of their causes, would try to compensate for his ignorance by inventing hypotheses about them [...]  For one who has no comprehension of physical law, but is aware of his own consciousness and volition, the natural question to ask is not: "What is causing it?", but rather: "Who is causing it?"  [...] The error [Mind Projection Fallacy] occurs in two complementary forms, which we might indicate thus: (a) My own imagination => Real Property of Nature, and (b) My own ignorance => Nature is indeterminate.

The philosophical dierence between conventional probability theory and probability theory as logic is that the former allows only sampling distributions, interprets them as physically real frequencies of "random variables", and rejects the notion of probability of an hypothesis as being meaningless. We take just the opposite position: that the probability of an hypothesis is the fundamental, necessary ingredient in all inference, and the notion of "randomness" is a red herring, at best irrelevant. [...] by "probability theory as logic" we mean nothing more than applying the standard product and sum rules of probability theory to whatever propositions are of interest in our problem.

We do not seek to explain "statistical behavior" because there is no such thing; what we see in Nature is physical behavior, which does not conflict in any way with deterministic physical law.

[...] as in any other problem of inference, we never ask, "Which quantities are random?" The relevant question is: "Which quantities are known, and which are unknown?"