Coffeehouse Post

Single Post Permalink

View Thread: Are computers becoming more human, or are humans becoming more like computers?
  • User profile image

    Manip wrote:
    Isn't it theoretically possible to build a non-deterministic structure on a deterministic one (the hardware). Although this kind of cuts the entire binary mind concept out, it doesn't make using a computer as a mind impossible. What about 'random' number generators for instance. Although it is difficult to make truly random ones as CPU speeds increase there numbers and the technology should produce MORE random results.

    Anything that your computer does that has the appearance of being non-deterministic will invariably be the result of external (or simply unknown) state being consumed. If the scope of the system you are describing is 'computer' or 'hardware' then you might consider this to be non-deterministic, because within your scope you won't have the information you need to be able to state that the outcome will be consistent regardless of the fact that all state within the scope of the system is in the same configuration (or more simply you may not know what that state is even if it is bound in the scope, classic example being the system clock). The only reason you can't make this claim is because inside your scope you can have no knowledge of the external state that will impact the system.

    Again, I want to stress my point that I don't like the word deterministic. The problem is that everyone uses it to imply 'not random' or 'we can know what the result will be', but this is not true. The definition I have for deterministic is "an inevitable consequence of antecedent sufficient causes", which does not imply anything about the need to be able to predetermine those inevitable consequences. It's not my fault that people hear big words and start using them (I too am regularly guilty as charged, and I suppose we should pride ourselves on our ability to learn from the results of engaging in experimental activity. It is after all this ability that is really at the heart of this thread (before I hijacked it to preach about determinism)).

    I have the same complaint with the word 'fate' btw, which has equally been watered down through popular romantic use such that it now carries all sorts of unwanted connotations. The definition I have for fate is "an event (or a course of events) that will inevitably happen in the future".

    There are any number of things that make your computer so complex that it has the guise of being non-deterministic. But it draws on all manner of unknown internal state (timers, counters, the present value of the fifth field in row four of table A, which sector the last blob of data got dumped to the page file on the disk, etc.), external state (electrical, thermal, physical (i.e. tangible), etc.) and is of itself only the result of very crude configurations of physical stuff (like copper and silicon) performing crude manipulations of stuff also bound to physical reality (like electrons, electromagnetic interference, etc.).

    I defined the universe as everything that can impact me. This means that I exclude the possibility of a system that can have any dependency on state that is not contained by the system. I did not exclude the possibility for other systems, but deny that there is any ability to receive state from them (otherwise they would be 'in scope').

    Before someone throws keywords at me like "Heisenberg's uncertainty principle" or "Chaos theory" please do me a favour and read about them. They only comment on the ability to configure state, observe state, or predetermine resulting state. This is about sensitivity to initial conditions which, based on current (solid) theories, can not be known or recreated.

    Something that is deterministic, can not by virtue of the fact that it is deterministic, render results that are non-deterministic without constraining the scope in which you define determinism such that you can rely on unknown state. Since that is what I believe, I deny that the universe contains anything that is non-deterministic. I don't deny that I can limit my scope and claim non-determinism, knowing that if I expanded my scope I would find determinism.

    Anyway, as a result of this conversation I realised something this evening. If it was indeed possible for the universe to arrive in an exact state that it had already been in, assuming the rules were constant (and really they must be, otherwise they are themselves state) then the universe would be stuck in an infinite loop! I thought that was a cool thought.

    Anyone else feel like standing up and being counted as a believer in a deterministic universe? I've heard people argue against it, but then again the only seemingly intelligent arguments pretty much discussed the ability to predetermine state, or the possibility for loosing and gaining state. But if the universe can 'loose or gain' state then I say that the scope of the universe has been too narrowly defined, and reiterate that determinism is not pre-determinism, and does not require an understanding of the rules, simply the observation that with respect to your frame of reference there is only one past that is effecting you.