On Consciousness

How might artificial intelligence achieve consciousness?

It has been said: “Machines will become conscious when they start to set their own goals and act according to these goals rather than do what they were programmed to do”. Setting its own goal instead of following a program would be seen as a bug, a flaw. Perhaps in order for a machine to be conscious it must be flawed. Evolution happens because of flaws in DNA. Mathematics requires Gödel.

In order to be conscious, intelligence must be flawed; without flaws it is just a program. Without flaws in nucleic acids, no life more complex than a virus could ever have evolved. Gödel’s theorem states that any self-consistent system of mathematics must be finite. Without the flaw of original sin, we would be God’s robots. Without random errors there could be no life – the Universe would be a lifeless, deterministic machine. Attempts to create a society which is flawless always fail because such a society can never renew itself.


A Return to Empiricism

Chapter 16 of The Fluid Catastrophe by John Reid


10 Replies to “On Consciousness”

  1. The semantics is around consciousness, self awareness and free will. I am not sure we understand the question well enough just yet. As you know John, I have an insight into what I think can be self awareness in an AI system, at least like in the sense of a Turing Test, that you would not know it was a non living entity displaying the characteristic.

    When I woke up with this insight I could not sleep for days I was so excited and sure as I still am that it is how self awareness works in us, and how to replicate it in a machine.

    I agree entirely about the need for flaws, else I would not be here. So Mim is right after all, I am flawed.

  2. Wow !
    This is a generalization, a higher order of,
    a better way of thinking about my idea about
    how humans diverged from other primates :
    An instability forming in the Theory of Mind cognition part of the primate brain,
    its having survival value, by making us more creative, bolder and so on, with the odd mad individual as collateral damage.

  3. My view is that the only facets of human conciousness that are not accessible to AI are Art, aesthetics (perhaps), spirituality, and philosophical questions such as the search for meaning. Questions like “do I still exist if my battery goes flat?”
    That is not to say that these proclivities could not installed to create an impression of human-like consciousness. Who can forget HAL in 2001?

  4. The original error would have been a Quantum fluctuation . So too most of the others . . . in what would become Life, ultimately what I have long been thinking of as a Universal Life-form, an extension of Life throughout the Universe, my idea of God.
    Also thus explained are various paranormal phenomena via Quantum Entanglement and so on, Life being thus universal and Quantum Mechanical. Hence Life’s being so weird . . . Brexit, Trump, present-day PC-madness and so on.
    https://en.wikipedia.org › wiki › Quantum_fluctuation
    In quantum physics, a quantum fluctuation (or vacuum state fluctuation or vacuum fluctuation) is the temporary change in the amount of energy in a point in space, as explained in Werner Heisenberg’s uncertainty principle. This allows the creation of particle-antiparticle pairs of virtual particles.

  5. So the answer to the Question :
    How might artificial intelligence achieve consciousness ?

    is that the Universal Life-form (ULF) has long been across Earthly AI developments, Gödel and so on. So, having long civilized Quantum Space via tweaking Quantum fluctuations and so on, ULF will be ready and able when our AI has reached a certain level, to tweak it, spark it into life via a para-normally-delivered Gödelian package suited to the purpose.

  6. What a great discussion. I particularly like Alan’s comments the we don’t understand the question well enough yet and that we need something like a Turing test. Can you spell out tour insight in more detail, Alan?

  7. It has been a good discussion.

    I’d love to explain more, but then I’d have to quantum fluctuate you all.

    Actually I am desperate to document it, because I think it is important, in the sense od=f David Marr’s vision efforts, that if it is not the answer it may be part of the answer and get us thinking on the right different track. My insights are not the whole answer but the bit it seems to address is very thought provoking.

    My plate is full at the moment and as soon as it is cleaned up, I’ll do something on this. I wonder how to categorise the fact that I think about consciousness? Is that consciousness contemplating consciousness? A tautology.

    Sir Roger Penrose’s “quantum nature of consciousness” will interest some readers. Not my idea though.

  8. An interesting discussion!
    It seems to me that there is an unproven assumption that life evolved due to flaws in chemicals.

    My theorem is that Self replicating machines are a result of, a symptom of intelligence. Lathes and computers spring to mind as machines which require themselves for replication. (well almost! I know how to make either without needing an original, but it would not be obvious how to do that to caveman coming cross these items for the first time.)

    In life there are three levels of self replication:
    1. Multicelled organisms need themselves to replicate (man begets man)
    2. Individual cells are created by mitosis – the splitting off of a copy.
    3. And very tellingly, RNA requires RNA for its creation: The mechanism is circular: Specific proteins translate DNA into RNA. The RNA and other proteins make the biological machines which make all the proteins .

    All three of these are exceedingly complex. It is very difficult to see how any of them, but particularly the molecular level one, could have evolved. The odds against even the simplest working system spontaneously springing into life are astronomical.

    Similar to the caveman looking at the complexity of my laptop.

  9. Hello everyone. At the 2001 conference CONSCIOUSNESS AND ITS PLACE IN NATURE: TOWARD A SCIENCE OF CONSCIOUSNESS, held at the University of Skövde, Sweden I gave a paper called


    with the

    ABSTRACT: Rational automata, implemented by the author and co-workers, engage in discourse with one another in correct English. Their conversation appears natural. They reason inductively and deductively, learn, remember and forget, and display human-like characteristics, such as curiosity, gullibility and duplicity. I surmise that rational automata will pass an OBJECTIVE Turing test — with the judge as passive observer, rather than participant. Rational automata, increasingly, will appear to be conscious.

    I’ll be pleased to send a copy of that paper to anyone who asks for it by emailing dfsander@mac .com or phoning 04 35 50 90 65.

Comments are closed.