Essay Abstract

I do not know quantum theory, but I know this: The most unpredictable thing is the human mind. Or at least, I believe that it is. And I believe the implications of this are profound. We have been trying to model something unpredictable (the human mind) with something predictable (artificial intelligence).

Author Bio

It's complicated.

Download Essay PDF File

i Grace, you may find this video of Max Tegmark of interest. https://youtu.be/9atnfAHBfSI Can also be found in the blog "Will A.I. Take Over Physicists' Jobs? More on Max Tegmark at the 6th FQXi Meeting" I think it is relevant to your writing, Concerning making AI trustworthy.A very important issue. The potential for cruelty to AI in the future,if they are programmed to have, or develop emotions of some kind is also worrisome..

    "We have been trying to model something unpredictable (the human mind) with something predictable (artificial intelligence)." Nature is the sorcerer, and man is the sorcerer's apprentice. Do men deceive themselves in believing that they can predict and control intelligence?

    According to Leonardo da Vinci:

    L'uomo ha grande discorso del quale la più parte è vano e falso, li animali l'hanno piccolo ma è utile e vero; e meglio è la piccola certezza che la gran bugia. (Man has sophisticated communication, most of which is vain and false, the animals have less sophisticated communication, and better is the small certainty than the big lie.)

    "Leonardo da Vinci", Wikiquote (Italiano)

      The various things which AI can do nowadays are amazing; I liked Tegmark's intro where he outlined all the improvements in AI's abilities recently. The question of whether AI is trustworthy seems to fall into two categories: First, is it accurate? Which is what Max Tegmark talks about. And it seems to be getting scarily accurate, especially if they incorporate all of the elements he outlined. But the second category is, is it altruistic? Which is a whole other ballgame. I suppose we have to decide, do we WANT to give AI consciousness, free will, or at least the appearance of consciousness? Because if we do so, then we have to consider not just its accuracy, but its altruism.

      If I understand correctly, you're saying that, if we think we can create intelligence--create AI with consciousness--we're deceiving ourselves. Which may be true. I do not think we can give a machine life. But consider the Turing test, which measures whether a machine is indistinguishable from a human in intelligence. I think we can achieve that. And I think we can create one which is indistinguishable from a human in emotion, as well. I do not think we can create life. But we can create the appearance of it.

      If we achieve that, it is going to be really, really confusing for those trying to decide if humans have souls--if these newfangled computers just "seem" to have souls, can't that be true of humans too? And I think the answer to that is simple. It's not "I think, therefore I am," it's "I am, therefore I am." Which, incidentally, can be taken two ways. First, because we recognize our own consciousness, we can extrapolate to other humans. And secondly, perhaps physics cannot answer the question of whether we have a soul--maybe we should just leave that to religion. Because if you believe in God, then having a soul is a logical conclusion.

      One interesting question would be, would a "conscious" AI treat other such AIs as though they possessed true consciousness? Would it "recognize" something in them, just as humans do in other humans?

      Dear Grace M Lo Porto, you write that you do not know quantum theory. I believe that nobody knows her. The level of abstract thinking in it rolls over to an inconceivable level. Her famous connoisseur Dirac told everyone who wanted to understand her "Shut up and count!" I made great efforts to make it clearer. The identity of space and matter of Descartes helped me in this. I made several discoveries after I realized that space moves, since it is matter. I set forth these discoveries in my essay: "The transformation of uncertainty into certainty. The relationship of the Lorentz factor with the probability density of states. And more from a new Cartesian generalization of modern physics. by Dizhechko Boris Semyonovich »

      New Cartesian thinking led me to the conclusion that our brain creates an image of the external world not in itself, but outside itself, in the space around our body. With the help of this image, our brain controls the movements of our body and can even cause levitation, since this image has a material basis. This is the difference between human consciousness and artificial intelligence.

      I invite you to discuss my essay, in which I show the successes of the neocartesian generalization of modern physics.

      Boris Dizhechko

      You say: And just like us, it is not to be trusted completely.

      May I comment this:

      If this implies that a human being can´t be trusted with specific issues, we will find out what such issues are. Those we will only find by trial and error, because in front of all experience in Time is the Unknown.

      Those issues will predictably found more, better and sooner with computers, regular and quantum and so with and in Human and Humanoids, by which correction can happen. That is in the evolution of Being. In that we can trust as there is always and all Times anyhow Being. That Being finds its best ways for Being. At least the best way at a given moment.

      Because of that we need room for trial and error and thus complete trust under surveillance by acquired and computed knowledge and understanding. Then also the human(oid) Mind, Consciousness, will be equal to Love combined with Care and Integrity, of which Care can be automated, Integrity monitored and Love felt. That I foresee and predict, and also decide to go that direction. Do you join?

      Bests Jos.

      13 days later

      Grace M Lo Porto free will vs consciousness

      The human mind is unpredictable. Trying to compare artificial intelligence with what you call 'free will' predictability is impossible due to infinity. Where does this free will come from? A person's brain is fed an infinite number of inputs. Even in early life nobody else has the same inputs. The extension of the infinite increases with age so "internal 'will'' options increase and distinctiveness, chance, and unpredictability increases. A machine similarly could be not be fed all the same inputs. The hindrances include the senses. Beyond that the inputs that exist every second cover wide ranges for sight sound smell etc. So the machine cant duplicate a particular human, but with enough inputs it could appear to be one human itself, just like us it would not be predictable. Its not complicated, it's infinity.

      Paul Schroeder

      17 days later

      Grace,

      For such a weighty mission, perhaps your brevity is warranted in the sense that maybe it says it all in its implications and a spurred imagery. But somehow I feel that it's like an unfinished symphony: not enough movements. It is certainly thought-provoking and a contribution to the community.

      And perhaps by implication you cover undecidability and uncomputability but give a lot of agency to unpredictability. But we still do have to program a quantum computer and therefore, it has human failings built into it. We do know, as I point out in my essay, a quantum computer can handle millions of variables at once rather than one but, as far as we know, cannot reroute neurons or build new connections as our brain does.

      Your warning, about limitless power, is certainly timely and subtle enough, but without programming it with truly deep learning, is it a danger. Or is that your point?

      Enjoyed your essay.

      Jim Hoover

        Thanks for your comment. It seems that, at its fundamental level, the universe is not predictable. Now, it also seems that, if something were static, it WOULD be predictable. A calculator, given the same inputs, will give the same outputs. A regular computer, with the same inputs, will give the same results, albeit with a few bugs and occasional crashes. These things are predictable, and they, by definition, do not change. So...could it then follow that, if something is not predictable, it is not static? Something like a quantum computer? I am not privy to the minute workings of quantum mechanics. But from what I have read and seen, the movement of fundamental particles, while it can be summarized and quantified, cannot be completely pinned down. We cannot say "If the universe does this, it will then do this," or "This particle will now move here," at least not with 100% certainty. And if we can't do that...if we can't get the same outputs from the same inputs, can we say for certain that it does not change? And can we say that a computer, based on those fundamental elements, would not change?

        I suppose the fundamental question is, why do we not understand quantum mechanics? Are our models just not good enough--is there some force we have not yet identified that would allow us to predict what happens with 100% certainty? Or, perhaps, is it that the universe, at its most basic level, is as changeable as the human mind?

        Oh, and one more thing--I enjoyed reading that you enjoyed my essay. So...thank you.

        a month later

        Grace,

        Wanted to let you know that I updated my essay and uploaded it a few minutes ago. Personally I feel that it is greatly improved. I did rate yours on 3/20.

        Please check mine out if you have time. Such honest, No BS, reviews are needed by all of us.

        Jim Hoover

        a month later

        very great essay. wish you typed on. wonderful message well passed on. hope AI moral fabric will be sustainable for us humans to evolve side by side. please read/rate on how Human bias may have made our world here https://fqxi.org/community/forum/topic/3525.all the best to you.

        8 days later

        Dear Grace M Lo Porto,

        You said.... But I will say one thing more...such a computer would be just like us. And just like us, it is not to be trusted completely........ Thats the concept told in the Arnold's movie "Terminator's" Skynet computer system. Well said.

        You also said you dont know quantum theory,

        For a similar philosophy (I also dont know much of quantum theory) in my cosmology essay " A properly deciding, Computing and Predicting new theory's Philosophy" , Hope you will have a look at it....

        Best

        =snp

        6 days later

        Grace,

        Hope you have time to check mine out before the deadline: https://fqxi.org/community/forum/topic/3396

        Jim Hoover.

        Hello Grace,

        A lovely short essay, but not a definitive statement of your premise. As it turns out; I agree wholeheartedly. Unfortunately; there is no longer a possibility to avoid giving AI too much information. We are left with the prospect of laying the groundwork to construct autonomous robots like R2-D2 and C3P0 or the possibility of being enslaved by Terminators. So the challenge is to imbue artificial intelligences with subtlety of judgment. And since they are not just working with 1s and 0s, quantum computers might offer a way to make computers able to think more subtly.

        I think we don't need or want to wait for that revolution to play out. We can already create computers with subjective search capabilities using non-associative algebras like the octonions. I am presently working on developing the foundations to make this possible. But it is a race against time. because brute-force AI is getting more and more powerful. It will certainly get to a point where artificial intelligences do think for themselves. But it is our job to assure that they do the job well, and are trustworthy or ethical.

        Warm regards,

        Jonathan

        Hi Grace - I enjoyed your essay. Short and sweet. And I agree that the human mind (and specifically the qualities of autonoetic consciousness) is the most unpredictable thing in the unviverse - and precisely because of free will.

        If you have a chance you might like my essay which goes into the special qualities of the mind in more detail. I do make some effort to characterize quantum physics (but cannot follow the math in spite of my math degree c.1973).

        Regards - George Gantz: The Door That Has No Key: https://fqxi.org/community/forum/topic/3494