Artificial Super Intelligence - are we ready for it?

[Stavros]

[QUOTE=trish;1618193]
A snow flake consists of nothing but H2O and yet each is unique. Why? Not because it has a soul, but because of the multitudinous variations physics allows in the formation of ice crystals and the multitudinous fluctuations in the ambient environment of these delicate lattices as they grow. So too with living creating creatures.
There are one hundred trillion neurons in a human brain. It would require two to the power of one hundred trillion bits of information just to describe with exactitude the physical state of that machine. Clearly we won’t ever be predicting human behavior on the level of neurons. To understand why humans sometimes write poetry we need higher level concepts. Perhaps ideas belonging to fields other than science. But clearly, with so many available biological variations and so much variation and fluctuation in our environment and our lives as we grow, mature and learn, it is natural to expect that when one human is found who writes poetry, not all will necessarily do so. Souls are not needed to understand the uniqueness of human beings. They may be needed to explain other things about humans, but I haven’t been told yet what that might be, or how such explanations work._

-I understand this argument, and I can see how powerful it is. Take, for example, babies who emerge from the womb without the clearly defined genitals humans ought to have, or who do not have all of their limbs or organs, or who are in some way -terrible expression I know but -'not perfect'. In many cases a genetic explanation will focus on the quality of the father's sperm, whether or not the man and woman were closely related, had medical or genetic problems of their own, and so forth. Science can as you suggest explain that genes are not all identical and that this will result in humans who are physically different from their parents.
However-

I’m not sure what “sum” means here. Surely it doesn’t mean the same as it does in arithmetic. Nor does it mean the same thing as “aggregate”, “collection” or “union.” We are not the mere collection of all of our parts, that’s for certain. But then, neither is a car engine. Take your car apart and put all the pieces into a huge box. It will no longer be your car. A car is not a simple sum of all of its parts. The parts of a car are integrated and interfaced in such a way that the state of the cars is intimately related by the laws of physics to the state of the car at later times.
-This I think is a weak argument, because two cars being driven off the production line must be exactly the same, so that the Ford I take possession of in say, Birmingham, is the same vehicle as the one John takes possession of in London. Just as I would expect an Apple Mac purchased in one shop to be the same as one purchased in another. Machines that have been designed down to the last rivet and chip surely cannot be subject to 'genetic' modification as humans can be?

The question of this thread is can human-designed, man-made machines achieve intelligence? sentience? and if so are they an existential danger to humans? You seem to think that no man-made machine will ever have a soul and so we will never have to grant such a machine our empathy, our sympathy, our respect etc.
-I don't see why a human should be upset if the kettle leaks, and cradle it as a consequence. Brutal as it sounds, I throw that kettle away and buy a new one. I don't do lullabies for vacuum cleaners. Can machines be designed that switch themselves on and make the tea just before you wake up? Yes of course, but it is still dependent on electricity, and on a human putting water and tea in the machine. Can computers be programmed to switch themselves on and perform functions? Yes, but at one point do they 'think for themselves'? To me this is science fiction, but also illogical. I don't see how a machine can evolve by itself, and do not think that AI can make the leap from human dependency to autonomy.

The hypothesis that thinking beings require divinely bestowed souls grants you a slam-dunk argument. I think it’s very unlikely that humans will ever create sentient machines, but I don’t have a slam-dunk argument that concludes no such thing is possible. Indeed I think our very existence illustrates the principle of a sentient physical system. So I can’t rule the possibility that we may want some day in the future to grant to some of our machines the same respect that we (ought to) afford others.
-This is where science and religion cannot meet. Science will argue that the child has a genetic disorder where a religious person will say it is karma, or God's Will. Science is satisfactory when it comes to the physical, but is still struggling to explain the autonomy of the person, of an identity that can be shaped in spite of their bodily reality, such as Michael, who on weekends is Michelle. But religion does not really explain it either, since the conclusion that a situation is 'God's Will' is to me a meaningless statement as I do not know how anyone can know God's Will, not least because this God seems to will things for one person that are opposed by another claiming the same authority.

As I suggested, we are not really making progress with this, although I do think it is an interesting thread.

[Stavros]

Stavros and Trish both love to LORD it over us with their engorged vocabularies, I think my only unfinished business here would be to hear Trish admit to at least the POSSIBILITY of a Universal God that starts to cook where our paths end.

I would not want to Lord it over you, I hope you agree to that. As for Gods' cooking, I think, on the evidence of the last 5 billions years, one might be inclined to say 'Too much salt, guv'nor.'

[trish]

Trish said,

I’m not sure what “sum” means here. Surely it doesn’t mean the same as it does in arithmetic. Nor does it mean the same thing as “aggregate”, “collection” or “union.” We are not the mere collection of all of our parts, that’s for certain. But then, neither is a car engine. Take your car apart and put all the pieces into a huge box. It will no longer be your car. A car is not a simple sum of all of its parts. The parts of a car are integrated and interfaced in such a way that the state of the cars is intimately related by the laws of physics to the state of the car at later times.

Stavros replied

-This I think is a weak argument, because two cars being driven off the production line must be exactly the same, so that the Ford I take possession of in say, Birmingham, is the same vehicle as the one John takes possession of in London. Just as I would expect an Apple Mac purchased in one shop to be the same as one purchased in another. Machines that have been designed down to the last rivet and chip surely cannot be subject to 'genetic' modification as humans can be?

It would be a weak argument were its aim to prove we cannot make machines that behave predictably (at least within a reasonable degree of tolerance) in predictable situations; but the example was put forward to illustrate that even cars are not just a simple sum of their parts, thus undermining the position that humans must have souls because each human is more than the sum of her parts.

Let me paraphrase the greater-than-argument. It goes like this. You are more than the sum of your parts. The extra bit (the difference between you and the sum of your parts) must be your soul.

The argument is clearly not meant to apply to cars. To avoid such an application, the premise must interpret the word “sum” in a crucially different way and in doing so, it (the premise) assumes the conclusion (you have a soul) rather than proves it. This logical fallacy is called, “begging the question.”

-I don't see why a human should be upset if the kettle leaks, and cradle it as a consequence. Brutal as it sounds, I throw that kettle away and buy a new one. I don't do lullabies for vacuum cleaners. Can machines be designed that switch themselves on and make the tea just before you wake up? Yes of course, but it is still dependent on electricity, and on a human putting water and tea in the machine. Can computers be programmed to switch themselves on and perform functions? Yes, but at one point do they 'think for themselves'? To me this is science fiction, but also illogical.

Of course I’m not arguing that kettles and vacuum cleaners are sentient; only that some machines are (e.g. us) and that there is a possibility that humans might someday craft sentient machines (not that I think they actually will do so, especially anytime soon).

I don't see how a machine can evolve by itself, and do not think that AI can make the leap from human dependency to autonomy.

Nothing “evolves” by itself; but is induced to evolve by way of myriad of interactions with a complex and chaotic environment. A single human being modifies her outlook on the world and her responses to it not because she is possessed by a divine spirit or is in possession of a soul, but because she interacts with and is influenced by the world around her.

-This is where science and religion cannot meet. Science will argue that the child has a genetic disorder where a religious person will say it is karma, or God's Will. Science is satisfactory when it comes to the physical, but is still struggling to explain the autonomy of the person, of an identity that can be shaped in spite of their bodily reality, such as Michael, who on weekends is Michelle. But religion does not really explain it either, since the conclusion that a situation is 'God's Will' is to me a meaningless statement as I do not know how anyone can know God's Will, not least because this God seems to will things for one person that are opposed by another claiming the same authority.

Let me reiterate that my views in this thread shouldn’t be taken as those of Science. I think we agree that neither science nor “God’s will” satisfactorily explains the phenomena of sentience nor the apparent autonomy of persons. I would extend this judgment of explanatory failure to the soul-hypothesis as well. Given the state of our scientific, theological and philosophical knowledge, shouldn’t we leave open the possibility that some machines might be sentient by virtue of the physical integration of their parts and their complex interaction with an enormous, chaotic and hugely varied world?

[Stavros]

Of course I’m not arguing that kettles and vacuum cleaners are sentient; only that some machines are (e.g. us) and that there is a possibility that humans might someday craft sentient machines (not that I think they actually will do so, especially anytime soon).

I actually agree with a lot of what you propose, but not the quote above, I just find it too cold (too soulless?) to be described as a machine. It is as cold as that quote attributed to Stalin: the death of an individual is a tragedy, the death of a million is a statistic.
Not a scientific response, I know, but probably the cause of the doubts I have that AI/man-made machines will evolve by themselves. I don't know if it is because I grew up in a world where most people did not have a tv, where computing was rare and robots associated with the B films we saw at Saturday Morning Pictures. But even with the profound changes to AI, I still can't see beyond the buttons and lights. Maybe it's just my age.

[trish]

Let me be clear. I’m not saying sentience is a matter of computation. The modern theory of computation is a branch of mathematics. Not too long ago it went by the name Recursion Theory. It was founded by Turing, Godel, Church, Kleene and others as a sub-branch of mathematical logic within the field of pure mathematics. Were sentience simply a computational matter, an emergent property of a class of Turing machines (or their equivalent), then the study of consciousness would be reduced to a branch of pure mathematics. There would be no experiments to perform; just definitions to delineate and theorems to prove. For no good reason, my intuition runs counter to this. I do not think mathematics alone can encompass what we call sentience.

I think of sentience as a natural phenomenon. To me the world is not cold. It is complicated; complex to the point of being incalculable. It is a web of difficult and incomprehensible things: matter, energy, spacetime, fields and all the things embedded within and constituted by these things; all interacting in quantum bizarre and geometrically convoluted ways. Mathematics is cold and abstract. The world of things is hot, vast and filled with possibility.

Perhaps our difference just lies in what we understand a machine to be. For me, just about anything in the natural world that transitions from one state to the next according to natural laws as it interacts with the world is a machine. I’m inclined to the intuition that everything in the world belongs to the world (i.e. is natural). You, perhaps, are inclined to the intuition that some things (souls in particular) do not (i.e. some things are supernatural).

Perhaps we simply disagree on what “natural” means. I’m inclined to think that should it be proven that humans have souls (I’m not holding my breath), then through that proof we will have some footing toward working out how souls interact with their hosts, with each other and the rest of the world. We will begin to discover how they fit and function in the natural world; i.e. souls themselves would be understood, not as supernatural, but natural entities that belong to the world and arise from the natural world. Your inclination may be to draw a line and divide existence between the natural and the unnatural. Mine is to erase the line and let nature encompass all.

Even allowing for these different perspectives, our disagreement on the existence of souls currently remains a substantial one.

[martin48]

So we do not understand fully how our brains work, we do not understand fully the influence of nature and nurture, but we can observe the amazing similarities of identical twins who are raised apart. Nature – our genetic base – probably has more influence that we think.

So individuals are different in behaviour and beliefs; that is understandable. We do exhibit, or we believe we do, freewill.

If there are gaps in our observable knowledge then why jump to filling in this gaps with something called the soul? What is our evidence? It is a seemingly useful handle to cover our ignorance. But that is exactly what it is, something that keeps us ignorant.

[fred41]

A person is given a genetic blueprint and so a starting course is already written out...but prenatal and postnatal environmental effects further shape a person whose every decision during his/her early life can alter his/her personality and who they 'are'.
A person with an identical twin with developed schizophrenia has, if I read it correctly, a 48% chance of also having the disease based on genetics. Compared to the average person this is quite high, but note that it also means the person has a more than half a chance of not getting the sickness (or activating it). Doctors believe prenatal conditions, early childhood diseases, high levels of stress, substance abuse, avoiding social interactions...in other words - environmental factors, all play a strong role in whether or not a person with a predisposition to the illness will actually develop it. My point is that - how a person develops emotionally can be tweaked in so many nuanced ways, by so many factors that the explanation of a 'soul' is almost entirely unnecessary. How the mind and body develop is complex because living beings are complex. (BTW do identical twins share the same soul that was split?)
There is no proof that an actual souls exists. It's just another one of those things that no one ever saw, felt, heard or tasted...but someone first came up with the idea, somewhere in time and now people continue to believe in it based on faith. Sure, science can't disprove the existence of a soul...but it can't disprove the existence of pixies either...they're just real good at hiding.
There is no war between 'science' and religion. Science is a tool, a continuous gathering of information. Religion is faith in the supernatural...it's spiritual.
But people will always believe what they want to believe...they prefer romance to facts. It's why Uri Geller still cons people even though James Randi has shown him to be a fraud.
We don't understand everything in life and we may never completely understand life itself. To that end I guess you can use the word "soul" to mean - the thing that gives a person the spark of life, much as Victor Frankenstein's monster used electricity...the energy that activated him.
But that's not the same as defining the word soul as a spiritual mass of feeling and personality...that is, in fact, the 'real' person, with or without flesh.
The earth is an incredible place when you think about it...even in that 'cold' scientific way. Everything plays a part in everything else....and it's incredible how cells, organisms, beings adapt and change as is necessary for life...souls or not. God or not.
I no longer believe in a monotheistic God.
but I still feel wonder when I see a deer or a family of foxes...or a shooting star...or hold someones hand. I don't need the added romance or poetry of faith in a higher being. What's in front of me in this life is good enough for me.

[fred41]

Went a bit off track there. To bring it back in context of the thread, I would say, I don't have a problem labeling the power that an artificial sentient being uses to stay 'alive' being labeled a 'soul'. But a consciousness would be something different all together.

[Stavros]

By way of responding to the various posts above, I should apologise for not always being coherent on this subject. It is something I have occasionally thought about, but probably because I shall not live long enough to see the quantum leaps in computing/AI that we are promised, I tend to think about AI as a social issue, such as the mechanisation removing humans from the production of commodities, and the challenge this poses for the state, so that it is more political than scientific.

I tend therefore to think of this by using cars and computers as examples: if I buy two Apple laptops I expect both of them to function in the same way, as man-made machines designed to be identical in every way. Trish makes the valid point that while humans can be viewed as machines with the same working parts, in reality, just as two snowflakes made by the same process look different, the multiplicity of neurons and other components to a human mean that like snowflakes we will be simultaneously the same but different, just as apparently, twins can exhibit remarkable duplications of thought feeling and behaviour, even if in some other aspects they retain a degree of individuality. One notes, as an aside, that in some ancient cultures, twins were considered a curse or a calamity to the extent that one of the two might be killed at birth. Rene Girard in Violence and the Sacred (1972, page 56) uses this in his discussion of mimesis as both a building block of human societies but also one of its potential weaknesses, as it leads to envy, covetousness and its expression in violence.

The argument in science that the 'soul' does not exist because observer-dependent science has not found it has been challenged by some scientists. Thus, in an article in Psychology Today Robert Lanza points out that 'weirdness' is as much a part of quantum theory as rationality, thus:
While neuroscience has made tremendous progress illuminating the functioning of the brain, why we have a subjective experience remains mysterious. The problem of the soul lies exactly here, in understanding the nature of the self, the "I" in existence that feels and lives life.

Scientists do rely on probablity rather than evidence as an explanation for phenomena that they know is happening but which cannot be seen, yet this is not considered irrational or 'unscientific'. Whether or not Lanza is stretching the boundaries to include 'the soul' in science you can judge for yourself here:
https://www.psychologytoday.com/blog...dence-says-yes

The social level at which this becomes interesting to me relates to the issue of social change and what Marx called the 'means of production' and the 'social relations of production'. Marx relied to a great extent on a flawed processional view of history which begins with primitive communism, and moves through revolutionary phases to feudalism, to capitalism (mercantile capitalism followed by industrial capitalism and for some later thinkers monopoly capitalism etc) to socialism and ultimately to communism, thereby reproducing in material terms Hegel's concept of consciousness as something that begins as nothing and through multiple stages of challenge and change matures and grows until it expands to a state of absolute consciousness that has been described as 'Hegel's journey toward the sunlight', but which might in other terms be Nirvana.

Crucially, what Marx attempted to do in volumes 1 and 2 of Capital was to show how human beings who had at one time made their own tools, farmed the land for their food, made their own clothes from animals and crops, etc, find themselves in capitalist societies where their tools and their expertise has been transformed into a significantly more productive machine, to which they have become merely an appendage, required to push buttons or perform the same menial task a million times in a 16 hour day. Marx believed this mode of production -in his case, factory production- and the social relations in which it took place, created a form of reification in which the relationship of people takes on the appearance of a relationship between things, or to put it another way, human communities defined by human identity are replaced by networks of monetized linkages. Marx believed Hegel had consciousness upside-down, and that rather than seeing everything as a product of the mind, Marx saw the material world as the source of consciousness and thus argued that the dehumanization of the worker in a factory was possible because the worker's consciousness of himself as a free person had been crushed, he became a wage-slave -but that by bringing a collective of workers together in one place, a 'working class' or 'collective consciousness' became possible which, if becoming political, could revolutionize both capitalist production and social relations, and push human society out of capitalism and into a wholly new experience of life.

Curiously, Marx also seems to see humans as machines, in the sense that the working class is forced to do the same mechanical things all day and every day, and also sees collective action as the source of hope for revolutionary change, where history suggests that the kind of revolution Marx advocated has ended up creating societies where individual identity is considered such a threat to the organization and survival of the state that such people are physically removed either through murder, or by sending them to the Gulag.

But does not religion also impose collective identity on individuals, and go further to argue that to be a Christian, a Jew, a Muslim etc, does not just mean believing the same things, but behaving in the same way? Is this a way in which human beings are again presumed to be 'merely machines' by movements which claim to have identified the author of the body and the soul? Religions may not describe their believers as machines, but seem to think they should behave in an identical or regimented manner -to attend Synagogue or Mass, or Friday prayers, to fast at a particular time and so on. To believe man and woman are the 'natural' order of things and that any other variation is a violation of 'God's law'.

So that in practice, science and religion actually converge on a wide range of beliefs and practices. And for the most part, it seems to be either a denial of diversity and individuality -communism, religion- or a denial that the self even exists other than as the behaviour of molecules and neurons in the brain transformed into language.

From this point of view, the soul need not be a separate thing from the body, but the means whereby an individual exercises the reflexivity of thought and feeling that enables that individual to make decisions -practical, moral etc- some of which are determined by the body, some by society, but some by an unseen probability which enables us to identify one person as different from another, just as it enables two men, William Shakespeare and Christopher Marlowe, to write at the same time indeed, on the same day, but with different degrees of literary and theatrical skill.

Taken into AI, the only way a machine could become 'sentient' would be if a human were to interfere with a design whose intention is to limit a machine's functions and give it the power to expand into an autonomous AI 'creature', just as in literature and film robots who rebel, or 'resurrected dinosaurs' who become carnivorous have been made that way, by mad scientists or through greed. One notes, as an aside, that Robots don't tend to be interested in world peace or love and compassion, but that must be a reflection of their creators prejudice.

But in a world which wants to re-clone the woolly Mammoth, re-introduce wolves into Britain (no thanks!), maybe the ultimate question is not, can we trust machines, but can we trust humans? After all, we have invented nuclear weapons, and continue to develop them, as if we were sleepwalking into our own destruction.

Finally, given that for some the soul is the proof of eternal life, is it not the case that a computer could in theory function for an eternity -at least as long as the sun shines?

[buttslinger]

ATTENTION SOULESS BASTARDS!!!

Greetings and Salutations to you all!

Religion is the opiate of the people-K Marx

INDEED


If only one person who ever walked the face of the earth was a witness to GOD, the God exists. Period.
Not the God who left you high and dry when you really NEEDED that bicycle in third grade, not the God that people believe in but don't understand, THE GOD.
I am not saying that everyone should run off to a Zen Monastery, I'm saying if you spent eight years of your life training to be a Marathon runner, and you achieved your goal by winning the Olympics, then all that stupid training and sacrifice might have been worth the effort and time spent. To you. The fact that the losers in the stands around you couldn't fully understand it wouldn't mean much, but you probably would wish that they could understand what you were feeling.
Even people that have seen God don't understand it, It surpasses understanding. Just because you can't put it in a test tube ,,,blah blah blah.

I am ancient enough to remember reading that egg farmers had some site on the internet that would instantly bring them every mention of the word "EGG" in the daily papers. (early google)
And reading that airplane designers didn't need wind tunnels because they could do those tests on a computer.
For me, that's pretty good, the fact that humans have their greasy fingerprints all over keyboards and internets is good. Computers enhance human intelligence.
Of course they also made it possible for some pricks to steal my Mom's IRS return.
Just like God is our Father, Computers are looking like they will be the father of our puny brains, the destinations our logic would go to if we had the ability to go that far and fast. So that's pretty exciting. I can't imagine any computer that wouldn't have guys plugged into it, and so far, yeah, Self Aware Computers exist only in the imaginations of science fiction writers. If some computer pushes the button that launches all the ICBMs in the American arsenal, I would still call that human error.

Even if by chance some electrical aura becomes self aware, I doubt it would destroy the human race or even care about the human race. We tend to be very foolish about our own importance.