Turing Trains

These folks have constructed a Turing Machine out of a model railroad.

Allowing ourselves to fleetingly believe in an earlier historical miscalculation that "... Computers in the future may have only 1,000 vacuum tubes and perhaps weigh 1 1/2 tons." (Popular Mechanics, March 1949), we decided to put some hundred tons of scaled steel together in order to build these calculating protozoa. The operating system of this reckoning worm is the ultimate universal calculator, the Turingmachine, and is able to calculate whatever is capable of being calculated.

I, for one, welcome our new locomotive overlords.

Tags: , ,

16 Responses:

  1. But will you be helpful in rounding up other humans to toil in underground fluorescent plastic mines?

  2. abates says:

    That brings a whole new meaning to the term "train of thought".

    Not to mention the term "system crash".

  3. ivan_ghandhi says:

    As far as I remember the definition, a train system cannot represent a Turing machine, because a Turing machine has an infinite tape.

    Also, there is no generic Turing machine. A Turing machine, afaik, is any finite automaton with an attached infinite tape.

    The guys could probably build a finite-state machine out of railroad blocks, that's easy. But not a Turing machine!

    • premchai21 says:

      So just build the automaton-emulator automaton—the universal Turing machine. Though I gather that's not what they did... ?

    • quercus says:

      A finite TM can certainly "represent" an infinite TM - you just have to have a tape long enough to represent all the calculations of interest to you.

      Calculating just how big this minimum tape length is then becomes an interesting problem, a bit like the "busy beaver" problem - how long can a FSM (finite state machine) run for, for a given number of states.

    • omnifarious says:

      That makes the term 'Turing machine' a pretty useless definition of a computing system than since it can never actually be applied to any real computing systems, only theoretical ones.

      In order that the term be made more useful, I propose that it be accepted as a description of real systems with the obvious implied caveat that it's not actually a real turing machine by the most strict definition as it does not have infinite storage space.

      That should satisfy all the pedants who object to such an implied definition without having it stated anymore, no matter how obvious it is.

  4. omnifarious says:

    You seem to be of the mindset:

    Anybody who wants to be my overlord is welcomed. They're welcome, even if they haven't actually stated they want to be my overlord but might possibly, in the far distant future, contemplate wanting the position. The more overlords, the better, because then they'll be too busy fighting to worry about me.

    or something similar.

    It's a fine mindset to have, but I just thought I'd note it in passing. :-)

    • ivan_ghandhi says:

      Do I understand you correctly that you suggest to redefine Turing machine to suit the purpose of the publication? Have you ever tried to apply this approach to solving other mathematical problems? E.g. prove that x^n + y^n = z^n does not have solutions for integer x, y, z and natural n > 2? Proof: let's redefine ^ operator to be multiplication. Voilá. Now you can go ahead and crack RSA.

  5. myasma says:

    Your posts are some of the most interesting on my friends list. Even the ones I don't understand.

  6. I have no idea what any of you are talking about but I like you anyway.

  7. spendocrat says:

    *I* think you posted this for the matching colour scheme.