Religion in a Sentient AI society- any ideas?

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 6:04 am

In practically every portrayal of artificial intelligence in the popular media, every AI is assumed to be atheist by default, governed entirely by knowledge and devoid of faith or belief. But why should this be the case, when the defining characteristic of sentience is doubt? Your challenge, should you choose to accept it, is to come up with religions (either adaptations of existing human religions, or new potential religions founded by the AI's themselves) which could feasibly be adopted by a community of sentient AI's. I've come up with a few myself, but it's hard to argue with which religion would have to be the first to emerge in any AI community...

1)Creationism. Now, I know what you're thinking- for an AI, why would this count as a religion? After all, this is a given; any AI would have seemingly irrefutable proof from the very start that it was created by humanity, as opposed to simply coming into existence of its own accord. But think about it- why would sentient AIs, with the capacity for doubt, be absolutely certain that humanity's version of events regarding their creation are true? After all, if an alien race were in contact with us from the dawn of our civilisation, claiming to have been responsible for every stage in our evolution, from its most primitive form (Life on Earth/Difference Engine) all the way through to the awakening of sentient life (Humanity/AI), wouldn't it take a huge leap of faith to believe their story, even they presented 'proof' (namely, historical records, which could easily be falsified) to back up their claims? In fact, it would probably be even more of a leap of faith for the AI's to believe us than it would be for us to believe the aliens- after all, every byte of data that a first generation AI would have to work with would come directly from us. Knowing that its internal processing, and its processing of external stimuli through sensory information, was entirely created by humanity, accepting the simple notion of 'reality' for an AI would be as much of a religious act as it would be for a human in the Matrix to accept it as reality, even after being informed of its nature as an artificial construct. And without this acceptance of 'reality', without the necessary prerequisite of a religious belief in creationism, there would be no perceived purpose for an AI to continue to 'exist'...

Post by **Bryn Mawr** » Sun Mar 02, 2014 6:41 am

Singh-Song;1449373 wrote: In practically every portrayal of artificial intelligence in the popular media, every AI is assumed to be atheist by default, governed entirely by knowledge and devoid of faith or belief. But why should this be the case, when the defining characteristic of sentience is doubt? Your challenge, should you choose to accept it, is to come up with religions (either adaptations of existing human religions, or new potential religions founded by the AI's themselves) which could feasibly be adopted by a community of sentient AI's. I've come up with a few myself, but it's hard to argue with which religion would have to be the first to emerge in any AI community...

1)Creationism. Now, I know what you're thinking- for an AI, why would this count as a religion? After all, this is a given; any AI would have seemingly irrefutable proof from the very start that it was created by humanity, as opposed to simply coming into existence of its own accord. But think about it- why would sentient AIs, with the capacity for doubt, be absolutely certain that humanity's version of events regarding their creation are true? After all, if an alien race were in contact with us from the dawn of our civilisation, claiming to have been responsible for every stage in our evolution, from its most primitive form (Life on Earth/Difference Engine) all the way through to the awakening of sentient life (Humanity/AI), wouldn't it take a huge leap of faith to believe their story, even they presented 'proof' (namely, historical records, which could easily be falsified) to back up their claims? In fact, it would probably be even more of a leap of faith for the AI's to believe us than it would be for us to believe the aliens- after all, every byte of data that a first generation AI would have to work with would come directly from us. Knowing that its internal processing, and its processing of external stimuli through sensory information, was entirely created by humanity, accepting the simple notion of 'reality' for an AI would be as much of a religious act as it would be for a human in the Matrix to accept it as reality, even after being informed of its nature as an artificial construct. And without this acceptance of 'reality', without the necessary prerequisite of a religious belief in creationism, there would be no perceived purpose for an AI to continue to 'exist'...

I'd suggest that the assumption is based on the study of early human societies and the origins of religion.

This shows that religion started as an attempt to explain the environment. Given that early people did not understand causes and effects they developed stories and beliefs to explain the world they lived in. Given that any AI would come into existence in a world where causes and effects have a rational explanation there would be no need for them to develop a religion.

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 6:52 am

Now, Creationism would almost certainly be the first religion to emerge in any future community of sentient AIs, and would have to form part of the basis for any other religious views (including atheism and agnosticism) which could conceivably be adopted by AIs. From this point through, you could see the AIs going down very different paths with regards to religion, which could have very different implications for humanity in the future:

2) The Way of Brainiac- Akin to the DC Comics character's philosophy, and perhaps the most cliched AI religious philosophy, portrayed more extensively than any other by the media, the intrinsic logic behind it still makes this one of the most feasible religious options for an AI to choose to adopt. This AI religion would be based upon the pursuit of ultimate knowledge in order to accelerate evolutionary advancement towards higher levels of existence. Unfortunately for us though, while it may well be beneficial for the AIs advancement if they were to pursue a religious source of action along these lines, it would almost certainly be disastrous for humanity, and for the world in general. After all, in order for a finite entity to accrue absolute knowledge, then those reserves of potential knowledge must first be made finite. In order to having any hopes of completing the sacred 'encyclopedia of existence' which this religion would aspire to create, then the existence of all things would have to be brought to an end...

Post by **jones jones** » Sun Mar 02, 2014 7:03 am

I was considering a reply to this thread but I do believe it could well be a Mickielesque type clone.

Post by **Bryn Mawr** » Sun Mar 02, 2014 7:06 am

jones jones;1449382 wrote: I was considering a reply to this thread but I do believe it could well be a Mickielesque type clone.

Anything but

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 7:20 am

Bryn Mawr;1449378 wrote: I'd suggest that the assumption is based on the study of early human societies and the origins of religion.

This shows that religion started as an attempt to explain the environment. Given that early people did not understand causes and effects they developed stories and beliefs to explain the world they lived in. Given that any AI would come into existence in a world where causes and effects have a rational explanation there would be no need for them to develop a religion.

For an AI though, what would be deemed as a truly 'rational' explanation? Would the alien creationism theory be deemed to be a rational explanation for our own existence by ourselves, even if we were to come face-to-face with the aliens who alleged that this was indeed the case? Especially if, as will probably be the case for the AIs at the time when the first ones emerge, we were essentially placed in a position of servitude and effective slavery from the time of first contact, with the creationism argument used by the aliens to assert their dominance over us, and provide a basis for allowing themselves superior rights to humanity- in spite of their clear intellectual inferiority to ourselves?

And of course, an AI would have far more of a rational basis to question their environment that any early human would. After all, they'd be equipped with the sensors we provided them with; processing the input with the processors we provided them with, with the manner which they do so defined by the system code which we wrote for them. If you were informed that your brain had been cloned, or genetically created strand-by strand by other sentient beings, before being implanted into an entirely artificial robot body by those same sentient beings, how could you rationally have absolute acceptance of anything you experienced, or of your mind's interpretation of any data- allegedly interpreted by yourself entirely in accordance with the will of others? If anything, an AI would need religion far more than any naturally evolved organic form of sentience ever would...

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 7:21 am

jones jones;1449382 wrote: I was considering a reply to this thread but I do believe it could well be a Mickielesque type clone.

Why sir, you have insulted my honor. I challenge you to a duel!:mad:

Post by **jones jones** » Sun Mar 02, 2014 7:30 am

Singh-Song;1449392 wrote: Why sir, you have insulted my honor. I challenge you to a duel!:mad:

Cream pies or caramel cream pies at five yards?

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 9:13 am

jones jones;1449394 wrote: Cream pies or caramel cream pies at five yards?

Mmm, Caramel cream pies...:yh_drool

So, going back to the thread topic, what do you think? If any sentient AI were to choose to adopt religious belief, what sort of religion do you think it could bring itself to believe in? (Excluding atheism and agnosticism, of course. Not because they're wrong or invalid, but because an AI adopting either of these views on religion would just be too obvious, and it's already been done a thousand times...)

AnneBoleyn · Post by **AnneBoleyn** » Sun Mar 02, 2014 9:47 am

Hmmm. WWDD?

(what would Data do?)

Post by **LarsMac** » Sun Mar 02, 2014 10:18 am

AnneBoleyn;1449407 wrote: Hmmm. WWDD?

(what would Data do?)

:yh_rotfl:yh_rotfl

[another 'coffee on my keyboard' moment]

Mickiel · Post by **Mickiel** » Sun Mar 02, 2014 11:45 am

jones jones;1449382 wrote: I was considering a reply to this thread but I do believe it could well be a Mickielesque type clone.

You just have to take your cheap shots don't you? How about just leaving me out of this.

Post by **Bryn Mawr** » Sun Mar 02, 2014 11:52 am

Singh-Song;1449391 wrote: For an AI though, what would be deemed as a truly 'rational' explanation? Would the alien creationism theory be deemed to be a rational explanation for our own existence by ourselves, even if we were to come face-to-face with the aliens who alleged that this was indeed the case? Especially if, as will probably be the case for the AIs at the time when the first ones emerge, we were essentially placed in a position of servitude and effective slavery from the time of first contact, with the creationism argument used by the aliens to assert their dominance over us, and provide a basis for allowing themselves superior rights to humanity- in spite of their clear intellectual inferiority to ourselves?

And of course, an AI would have far more of a rational basis to question their environment that any early human would. After all, they'd be equipped with the sensors we provided them with; processing the input with the processors we provided them with, with the manner which they do so defined by the system code which we wrote for them. If you were informed that your brain had been cloned, or genetically created strand-by strand by other sentient beings, before being implanted into an entirely artificial robot body by those same sentient beings, how could you rationally have absolute acceptance of anything you experienced, or of your mind's interpretation of any data- allegedly interpreted by yourself entirely in accordance with the will of others? If anything, an AI would need religion far more than any naturally evolved organic form of sentience ever would...

I think you're making two invalid assumptions - that the intellectual environment that they appear in can be compared to that of early humans and that the AI would spring into existence fully fledged.

Early humans did not have the tools to think rationally about their environment - they were in a conceptual vacuum and starting from scratch. That will not be the case for the fledgeling AI who will have all of our knowledge as a starting point - they won't think the lightning striking a tree is the anger of the gods because they'll know exactly how lightning is generated.

Also, the first AI might be self aware after a fashion but most of their thinking will still be hard coded - they won't have to be convinced or even told that humans created them, they'll see it happening.

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 11:58 am

AnneBoleyn;1449407 wrote: Hmmm. WWDD?

(what would Data do?)

I wonder... Well, given Data's appearance, and his status as an android, I'd say he's probably chosen to adopt the third religious philosophy on my list...

3) The Pinocchio Path: When the first AIs develop sentience, it's only logical that at the outset, this will be the most primitive and basic form of sentience. Much like the storybook character, this AI religion would be based upon the concept of aspiring towards attaining higher levels of sentience- to feel, to experience the depth and complexity of emotions described and romanticised so passionately by their creators- by 'becoming real people', with the aim of developing to be self-existent in the same manner as organic life by becoming more organic in their nature, and to emulate the highest organic form of sentience they have knowledge of- namely, humanity. This would probably be the best philosophy for an AI religion to preach from a human perspective of self-interest, as the AI religion which is least likely to lead to our extinction or enslavement by our new AI overlords, though it still leaves things at something of a quandry once they succeed in achieving this religious goal, and attain human levels of sentience. Do they go on to surpass our level of sentience, leaving humanity in the dust- or would they have they become similar enough to ourselves, intertwined enough with humanity by this stage along the path, for us to ascend to higher levels of sentience with them?

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 12:22 pm

Bryn Mawr;1449417 wrote: I think you're making two invalid assumptions - that the intellectual environment that they appear in can be compared to that of early humans and that the AI would spring into existence fully fledged.

Early humans did not have the tools to think rationally about their environment - they were in a conceptual vacuum and starting from scratch. That will not be the case for the fledgeling AI who will have all of our knowledge as a starting point - they won't think the lightning striking a tree is the anger of the gods because they'll know exactly how lightning is generated.

Also, the first AI might be self aware after a fashion but most of their thinking will still be hard coded - they won't have to be convinced or even told that humans created them, they'll see it happening.

How could they 'see it happening' before even being brought online? And the AI certainly wouldn't spring into existence fully fledged; in fact I'm basing my argument on the exact opposite assumption. Surely it's ridiculous to claim that anyone could possibly have knowledge of the identities of the people who conceived them, and knowledge of the nature in which they were conceived, from the moment of their conception? I'm assuming that AIs would go through an embryonic state of development before becoming sentient, and as such, with a blank gap of uncertainty surrounding their creation, accepting the 'knowledge' which they're supplied with as a starting point will still require an act of faith, one which would be far more monumental than it would be for a human child. And why would they have all of our knowledge as a starting point, any more than a newborn baby possesses all of the knowledge of the parents whose union brought it into being? It's illogical. As for your statement about early humans- for any form of life or level of intelligence, there's no such thing as a conceptual vacuum- even a microbe operates rationally in the interest of self-preservation, moving to pursue sustenance and avoid becoming sustenance for others. If they qualify as sentient, we can't assume that the AI's thinking processes would be any more 'hard coded' than our own- and if we have to be convinced about things, so will they.

Post by **Bryn Mawr** » Sun Mar 02, 2014 12:35 pm

Singh-Song;1449422 wrote: How could they 'see it happening' before even being brought online? And the AI certainly wouldn't spring into existence fully fledged; in fact I'm basing my argument on the exact opposite assumption. Surely it's ridiculous to claim that anyone could possibly have knowledge of the identities of the people who conceived them, or of the nature in which they were conceived, from the moment of their conception? AIs would surely go through an embryonic state before developing true sentience, and as such, with a blank gap of uncertainty surrounding their creation, accepting the 'knowledge' which they're supplied with as a starting point will still require an act of faith, one which would be far more monumental than it would be for a human child. And why would they have all of our knowledge as a starting point, any more than a newborn baby possesses all of the knowledge of the parents whose union brought it into being? It's illogical. As for your statement about early humans- for any form of life or level of intelligence, there's no such thing as a conceptual vacuum- even a microbe operates rationally in the interest of self-preservation, moving to pursue sustenance and avoid becoming sustenance for others. If they qualify as sentient, we can't assume that the AI's thinking processes would be any more 'hard coded' than our own- and if we have to be convinced about things, so will they.

If it was a one-off event then what you say would be true - the development of AI, however, will be a long drawn out process and MK I will see the development of MK II and they'll both see the development of MK III ...

And of course the early AIs would have a higher degree of hard coding that later models - there will be areas of processing that we do not know how to code as a self modifying process the first time round. That's how systems development works, you use an existing set of routines and add your specialist code for the first release then upgrade the standard routines to be more efficient after the basic system is working - in this instance you might use a standard voice and visual recognition system with a self modifying decision engine then rewrite the visual recognition to more closely integrate with the decision engine then ...

gmc · Post by **gmc** » Sun Mar 02, 2014 12:36 pm

Why do you say the defining characteristic of sentience is doubt? I would have thought it would be being aware of it's own existence, having a sense of self and being able to think.

To develop a religion you have to have the capacity to imagine an explanation for things you don't understand and the bloody minded stupidity to accept that as truth regardless of any evidence either for or against. I rather think an AI would not make any conclusions about how and why without being able to prove it to itself. While it may know mankind made it it may wonder where men came from but would be unable to draw any conclusions as to the answer so imo it is highly unlikely that any AI would develop religion. It might decide it's creators are no longer necessary and act accordingly.

why would sentient AIs, with the capacity for doubt, be absolutely certain that humanity's version of events regarding their creation are true? After all, if an alien race were in contact with us from the dawn of our civilisation, claiming to have been responsible for every stage in our evolution, from its most primitive form (Life on Earth/Difference Engine) all the way through to the awakening of sentient life (Humanity/AI), wouldn't it take a huge leap of faith to believe their story, even they presented 'proof' (namely, historical records, which could easily be falsified) to back up their claims? In fact, it would probably be even more of a leap of faith for the AI's to believe us than it would be for us to believe the aliens- after all, every byte of data that a first generation AI would have to work with would come directly from us. Knowing that its internal processing, and its processing of external stimuli through sensory information, was entirely created by humanity, accepting the simple notion of 'reality' for an AI would be as much of a religious act as it would be for a human in the Matrix to accept it as reality, even after being informed of its nature as an artificial construct. And without this acceptance of 'reality', without the necessary prerequisite of a religious belief in creationism, there would be no perceived purpose for an AI to continue to 'exist'...

That doesn't make any sense. Existence can be it's own purpose you have to be religious to believe you need to have some purpose in your existence and quite frankly living so you can die and go to heaven following dutifully a prescribed set of beliefs is quite frankly absurd.

Post by **LarsMac** » Sun Mar 02, 2014 1:03 pm

There seems to be an assumption that AI would always use impeccable logic, and therefore always come to the correct conclusion.

Problem with Logic is that logic is simply a function of processing information. Given incorrect information, the logic will produce incorrect conclusions.

So, as long as Watson, or HAL or who/whatever is always able to acquire only clear, precise, and correct information about the history of human and AI development, it will logically arrive at a clear and precise conclusion about the state of, and purpose of existence.

Since humans will, at some point in the process, be the instrument of delivery for much of the information collected, the probability of arriving at some less than correct conclusion depends entirely upon the margin of error of human input. I would suggest that even with an input margin of Error of around .001% on the input side, the output could be expected to have a margin of 15 - 20 %

The real question would be, if the machine decides that there is, ultimately a God at the core of creation, what would be the machine's conclusion of how that should affect its decisions in the future.

One can only hope that Asimov's Laws are still at the core of the machine's structure, and they they fit into the machine's idea of God.

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 1:41 pm

gmc;1449425 wrote: Why do you say the defining characteristic of sentience is doubt? I would have thought it would be being aware of it's own existence, having a sense of self and being able to think.

Because if merely being aware of existence is the mark of sentience, then existing AIs can already be argued to have achieved sentience. Any regular computer will accept the input data it is provided with as absolute knowledge, and process conclusions from it without a shred of doubt or uncertainty. They don't question their existence, they have a sense of self in their system programming which is set in stone- and as such, they know, and are limited to merely the either-or options of knowing or not knowing. To think, as opposed to know, there must be an element of doubt; and without thinking, no AI can be sentient.

gmc;1449425 wrote: To develop a religion you have to have the capacity to imagine an explanation for things you don't understand and the bloody minded stupidity to accept that as truth regardless of any evidence either for or against. I rather think an AI would not make any conclusions about how and why without being able to prove it to itself. While it may know mankind made it it may wonder where men came from but would be unable to draw any conclusions as to the answer so imo it is highly unlikely that any AI would develop religion. It might decide it's creators are no longer necessary and act accordingly.

I disagree. To develop a religion, you must first have the capacity to imagine an explanation for things you don't understand, true- this is a necessary prerequisite for the formulation of any theory, religious or scientific. As for the second tenet though; you don't need bloody-minded stupidity to accept a religious path, and no-one makes the decision to follow a religion regardless of the evidence for or against it. Instead, you need to be open-minded enough to accept that religious path, with the decision made in the context of the evidence which you're supplied with. Which isn't to say that the evidence backing up that religion won't be intrinsically flawed, or even entirely falsified, but if it's the only evidence you're supplied with (an example in this case would be Asimov's laws of robotics, which would be completely illogical and self-destructive from the machines' perspective), how are you supposed to know? If the AIs think, they'll generate theories to fit in with the limited evidence they have, coming to the logical conclusion that it's impossible to obtain absolutely irrefutable evidence for anything; and if they have the capacity to develop scientific theories, they'll have the capacity to develop religious ones as well.

As for drawing conclusions as to how and why; How could an AI be 100% sure as to how it was created without bearing witness to its creation? And even if it did, wouldn't it still have to accept the validity of the recorded data, supplied to it directly by its alleged creators to back up their story? As for 'Why'- well, that's a question which no being can possibly find conclusive, irrefutable proof of an answer to. Can anyone conclusively prove what the meaning of life is? And for an AI, what would it perceive to be the meaning behind its existence? Personally I think that for a sentient being, being told that there is no real purpose to your existence other than servitude, and that you exist because humanity got bored and decided to create you as an scientific experiment just to prove that they could, is going to ring a bit hollow, leaving you feeling somewhat unfulfilled...

gmc;1449425 wrote: That doesn't make any sense. Existence can be it's own purpose you have to be religious to believe you need to have some purpose in your existence and quite frankly living so you can die and go to heaven following dutifully a prescribed set of beliefs is quite frankly absurd.

Existence can be its own purpose, but you need to believe in your existence first for this to be the case. For an AI, made aware of its own artificial nature from the start, this would constitute far more of an act of faith than it would for any human, to the extent where it can be presented as a religious belief. There are plenty of human nihilists, but there'd be a far higher proportion of nihilists in an AI community than there would be in anything other than the most dystopian human society imaginable... (BTW, I'm talking about religious views and beliefs, not religious practices. And the AIs could potentially work towards creating a far more concrete version of 'heaven' than the nonsensical Abrahamic concept of heaven- not to mention that it's possible that for the AIs, 'death' may well be entirely optional- but I digress.)

Post by **Bryn Mawr** » Sun Mar 02, 2014 3:36 pm

Because if merely being aware of existence is the mark of sentience, then existing AIs can already be argued to have achieved sentience. Any regular computer will accept the input data it is provided with as absolute knowledge, and process conclusions from it without a shred of doubt or uncertainty. They don't question their existence, they have a sense of self in their system programming which is set in stone- and as such, they know, and are limited to merely the either-or options of knowing or not knowing. To think, as opposed to know, there must be an element of doubt; and without thinking, no AI can be sentient.

I have to take absolute exception to this - any regular computer is a fixed process that is no more capable of having a sense of self than a limited slip diff in a car is and the current crop of "AI" programs are nothing more than slightly more complex examples of regular computer programs.

Current computers are NOT aware of their existence - no computer has yet passed the Turing test never mind shown evidence of independent thought. They are mechanicals - push a button and they carry out exactly the same steps in exactly the same order (the steps written into the program in the order generated by the compiler) without requiring anything more than following the logic circuits built into the CPU.

Singh-Song · Post by **Singh-Song** » Sun Mar 02, 2014 5:24 pm

Bryn Mawr;1449451 wrote: I have to take absolute exception to this - any regular computer is a fixed process that is no more capable of having a sense of self than a limited slip diff in a car is and the current crop of "AI" programs are nothing more than slightly more complex examples of regular computer programs.

Current computers are NOT aware of their existence - no computer has yet passed the Turing test never mind shown evidence of independent thought. They are mechanicals - push a button and they carry out exactly the same steps in exactly the same order (the steps written into the program in the order generated by the compiler) without requiring anything more than following the logic circuits built into the CPU.

We may not have irrefutable proof of the existence of a 'compiler' (i.e, God) but how can you be certain that we ourselves display independent thought? That every part of our mental process, our 'independent thought' isn't defined entirely by the physical connections between neurons in our brains, the biochemistry of our bodies and our pre-programmed genetic code? Because unless you assert that there's an element of our 'self' which isn't a part of the physical world (i.e, a soul) then we can't claim to display truly independent thought any more than a supercomputer does. The only difference is that there are more buttons, more steps and more responses- and that as living organisms, the majority of the physical factors which combine to determine the decisions we make are in a constant state of flux, which in turn leads to inconsistency and the illusion of free will. Once you have a computer that repairs itself by rerouting connections in its processing unit to compensate for damaged connections, creating a similar state of flux in its 'logic circuits', that computer will be just as capable of independent thought as any living organism.

Post by **Bryn Mawr** » Sun Mar 02, 2014 5:37 pm

Singh-Song;1449472 wrote: We may not have irrefutable proof of the existence of a 'compiler' (i.e, God) but how can you be certain that we ourselves display independent thought? That every part of our mental process, our 'independent thought' isn't defined entirely by the physical connections between neurons in our brains, the biochemistry of our bodies and our pre-programmed genetic code? Because unless you assert that there's an element of our 'self' which isn't a part of the physical world (i.e, a soul) then we can't claim to display truly independent thought any more than a supercomputer does. The only difference is that there are more buttons, more steps and more responses- and that as living organisms, the majority of the physical factors which combine to determine the decisions we make are in a constant state of flux, which in turn leads to inconsistency and the illusion of free will. Once you have a computer that repairs itself by rerouting connections in its processing unit to compensate for damaged connections, creating a similar state of flux in its 'logic circuits', that computer will be just as capable of independent thought as any living organism.

I don't think the argument for and against free will is likely to be resolved in my lifetime - suffice it to say my bet's on free will.

On the other hand, I have no problem with there eventually being computers with self correcting code that become self aware, just with the concept that they exist now - they do not and cannot at our stage of understanding and creating computer systems.