The mathematics of artificial intelligence.

Discussion in 'Intelligence & Machines' started by TheFrogger, Feb 4, 2017.

  1. DaveC426913 Valued Senior Member

    Messages:
    18,935
    No, the problem is simply that collecting the raw materials of any complex working thing is the trivial part. Assembling them into that complex working thing is the lion's share of difficulty.
     
  2. Google AdSense Guest Advertisement



    to hide all adverts.
  3. Write4U Valued Senior Member

    Messages:
    20,069
    This would seem a daunting task, even as we are already using complex systems. In fact that might complicate the problem. It would be easier to use a system of gradual accretion of fundamental parts, which slowly grows ihe structure and potential when additional useful chemicals are encountered and absorbed.

    But now let us put identifiers at critical points of every item in the tray. IOW. one part can only attach itself to a specific other part (similar to left and right *handed* chemicals.at a simpler level.)
    Now shake the tray for a million years. There is now a probibilisty that eventually all complimentary parts will join and produce a functioning computer.

    I don't believe that anyone has suggested that this process was instantaneously, but might have taken millions of years for a single complex organism to emerge from the randomly scattered chemical interaction.
    Rome, as a completed complex structure was not built in One day. And so it is with the evolutionary process.
     
    Last edited: Feb 20, 2017
  4. Google AdSense Guest Advertisement



    to hide all adverts.
  5. Write4U Valued Senior Member

    Messages:
    20,069
    Really?, is H2O difficult to assemble? IMO, it was inevitable that out of 4 trllion, quadrillion, quadrillion, quadrillion chemical reactions during the life of just earth itself, the probability of even random formation of compound chemicals ranges from 99.99999% certainty to less than 1% certainty (rare events).
    Even so, as Hazen says, the enormous available size and time are bound to produce organic molecules and from there simple organisms may be fashioned. We are not talking laboratory environments, we are talking about the constantly *experimentation at a universal sacale. That changes the equation in favor of creation of living things.

    Consider the Cern collider/ Earth has one, capable of imitating what goes on trillions of times per second throughout the universe.
     
  6. Google AdSense Guest Advertisement



    to hide all adverts.
  7. DaveC426913 Valued Senior Member

    Messages:
    18,935
    Maybe, but you are losing site of the OP's question.

    He's not asking if we might use mathematical operators as a basic building block in the programming, engineering and construction of an AI, some day when we might understand neural systems well enough to try. He's asking if those basic building blocks will simply "produce" AI.
     
  8. DaveC426913 Valued Senior Member

    Messages:
    18,935
    The analogy to the OP's question is not "could these things assemble over 3 billion years"; it is "if we merely 'have' these things, will they produce life?"
    No, they won't.
     
  9. Write4U Valued Senior Member

    Messages:
    20,069
    It seems they did, albeit probabilistically.
    Some very qualified scientist are proposing that everything in the current universe is a result of a mathematical *Imperative*, given the size and scope and richness of fundamental elements created shortly after inflation and heavier elements during star formation. A mathematical process.

    An example would be gold which is created during a star going nova. Only recently have we been able to duplicate the process, but only in miniscule amounts using enormous amount of energy to duplicate the conditions which allow for gold to form (a rare event). Yet, if the earth has a fairly sizable supply of gold, just think of how many other planets have acquired gold.

    So it is with living organisms. If we can find life on an average rocky planet such as earth, why should life in some form not be present in planets of other star systems.
    To think otherwise seems hubris to me. Looking at the abumdance AND variety of life on earth, it is clear that life can express itself in myriad of ways, given enough time and a hospitable environment.

    As to the OP question, everything in the universes is a result of natural mathematical functions. This is why and how this universe came to exist in the first place. othing has changed in that respect. All physical objects in the universe are a result of mathematical functions. The human brain functions mathematically, AI would function mathematically. To my knowledge there is no mathematical prohibition to the creation of an artificial (pseudo) intelligence.
     
    Last edited by a moderator: Feb 28, 2017
  10. Write4U Valued Senior Member

    Messages:
    20,069
    Then it would not be Artificial, it would be Natural. We are the proof, but it took our assembly about 500,000 years on earth and there are many much older sentient species.
     
  11. DaveC426913 Valued Senior Member

    Messages:
    18,935
    Which is great, but it's a far cry from answering the OP's question.
     
  12. Write4U Valued Senior Member

    Messages:
    20,069
    >The mathematics of artificial intelligence.?
    I am no programmer, but in short; Binary processing of observed values,
    But as with all computer languages, binary processing of complex numbers and equations (pattern recognition) seems to present no problem
     
  13. DaveC426913 Valued Senior Member

    Messages:
    18,935
    Yes. Half a million years of human evolutionary biology is not related to the thread topic.
     
  14. Write4U Valued Senior Member

    Messages:
    20,069
    As an example of probabilistic evolution it is. But the human brain works by an electro-chemical function.
    Can we construct an AI that functions chemically as well as electronically?
     
  15. DaveC426913 Valued Senior Member

    Messages:
    18,935
    What does the topic have to do with probabilistic evolution?
     
  16. Write4U Valued Senior Member

    Messages:
    20,069
    AI is a planned and assembled 'organism" (artificial).
    But evolution has shown what is possible .
    "Natura artis magistra:"
     
  17. DaveC426913 Valued Senior Member

    Messages:
    18,935
    Right. Which has nothing to do with the thread topic.
     
  18. Write4U Valued Senior Member

    Messages:
    20,069
    Mathematical *operators* performing self-assembly? As far as I know mathematical operators have no value in and of themselves. No values, no results. The oprators are part of the mathematical functions of spacetime fabric, but they are functions, not values.

    C'mon Dave, I am trying to get past that useless question and am trying to find what would be necessary for our science to construct an autonomous AI.
     
    Last edited: Feb 23, 2017
  19. DaveC426913 Valued Senior Member

    Messages:
    18,935
    Exactly, which is the answer to the question.

    Yes. That's hijacking. Why not simply start a thread on that, rather than warping a poorly-formed thread to a new purpose?

    It's like building a castle on top of the remains of another castle that's sunk into a swamp. It inherits all the bad stuff from the first one.

    There is no price on solid land upon which to build all the castles you want.
     
  20. Write4U Valued Senior Member

    Messages:
    20,069
    If we are in agreement on this, why not continue to explore the topic?

    The OP Title is nowhere near that poetic landscape you paint.
    "The mathematics of artificial intelligence".
    There is nothing wrong with that OP title and leaves room for a variety of interpretations, without hijacking anything or purchasing castle ruins.

    It is the opening post which asks the wrong question and which has been answered by several people.
    I just offered what I thought might add to the discussion. There was no intent to hijack anything. My posts were intended to explore the OP Title, no more, no less.

    If I were to start a thread on this subject I would have called it "The mathematics of (artificial) intelligence" and we'd be right back to where we are now.
     
    Last edited by a moderator: Feb 28, 2017
  21. TheFrogger Banned Valued Senior Member

    Messages:
    2,175
    Hello again. Thank you all for your replies. Some interesting points raised.

    Firstly, we have not used the gold standard for years.

    Secondly we do not have to use simple operators alone: the values one to ten can be used, since humans use a decimal system (because we have ten fingers.)

    For example

    9
    8
    7
    6
    5
    4
    3
    2
    1
    0

    are used, then a column added to the left, and the symbols are RE-USED.

    hence

    Digit-Place
    9-10
    8-9
    7-8
    6-7
    5-6
    4-5
    3-4
    2-3
    1-2
    0-1

    Ten distinct digits (or values.)

    What about if I WROTE the following:

    1<A<10
    1<B<10
    1<C<10
    1<D<10
    1<E<10

    A+B*C-D/E
    A+B*C/D-E
    A+B-C*D/E
    A+B-C/D*E
    A+B/C-D*E
    A+B/C*D-E
    A*B+C-D/E
    A*B+C/D-E
    A*B-C+D/E
    A*B-C/D+E
    A*B/C+D-E
    A*B/C-D+E
    A-B+C*D/E
    A-B+C*D/E
    A-B*C+D/E
    A-B*C/D+E
    A-B/C+D*E
    A-B/C*D+E
    A/B+C*D-E
    A/B+C-D*E
    A/B*C+D-E
    A/B*C-D+E
    A/B-C+D*E
    A/B-C*D+E

    ...or put another way

    +=1, *=2,-=3,/=4,

    1*2*3*4

    As I stated, one of these permutations must produce some kind of intelligent result, if at all possible.

    However in the interest of thoroughness we must also include the equals sign, because it, too, is an operator. It may produce a WRONG answer, but it is an answer nonetheless. Again a question is raised: must a computer always be correct, or can a wrong answer enable learning...???
     
  22. DaveC426913 Valued Senior Member

    Messages:
    18,935
    Why?


    One would have to define "correct" and "wrong". They are contextual.

    Let me ask you if this equation is right or wrong:

    1+1=10
     
    Write4U likes this.
  23. Write4U Valued Senior Member

    Messages:
    20,069
    The eminently functional binary system , which, in context of quantum mechanics must precede all other mathematical functions. (on/off) = (cause<->effect) = (wave function - dynamic expansion)
     
    Last edited: Mar 24, 2017

Share This Page