Abhijit Naskar's Blog - Posts Tagged "thinking-computer"
Pretending to have sentience is not the same as having sentience: On Artificial Intelligence
The human mind is infinitely more complex than a computer and more importantly, it is born from the most magnificent organized organic structure in the known universe, that is, the human brain. The ultimate purpose of a computer is not to survive, but to compute. Even when it troubleshoots some of its internal errors to ensure sustained functionality, the proper algorithm for that entire troubleshooting process has to be embedded in the system by its programmer fist. Then comes the matter of awareness of being able to function.
On its own, a computer is not afraid to be turned off or malfunctional, but the human mind is, and this fear of death and ailment has been enabling our psyche throughout our days in the African Savanna to stay alert and avoid possible predatory attacks by all means. And we had been living with this sense of fearful alertness for so long, that substantial traces of it still remain in our neurobiology, which get activated when we watch horror movies – and that fear response gives us a certain amount of thrill, that pumps our body with adrenaline. This is the reason why many of us enjoy a good horror movie despite the fact that it may follow a sleep-less night or a sleep full of nightmares. We enjoy being on the edge because that’s how our primal ancestors spent their lives.
Now here is a question to ponder over – can we create a computer that can feel actual fear when there is an outbreak of a computer virus! The most we can do, is to install an anti-virus program into the system, so that if a known virus attacks it, the firewall will get activated and prevent any substantial harm to the system. But if it is a newly programmed virus, then the computer would become completely helpless, but again, it will not feel afraid or helpless, rather it will simply crash or stop functioning properly due to the infection.
Artificial intelligence is nowhere near attaining actual sentience or awareness. And without awareness it’s simply a mechanical device, which may pretend to show emotions and sentience, if it is programmed to do so, and thus it may be able to fool the humans as being alive, but in its own internal circuitry, it’d simply be following its preprogrammed tasks through the flowchart of an algorithm. At the current stage of our technological development, we can indeed create an artificial intelligence that can almost succeed in fooling the majority of the humans with its pre-programmed pretenses that it is sentient, but pretending to have sentience is not the same as showing signs of sentience. For example, Alexa or Siri may sound quite alive to many, but it’s simply following instructions on some algorithm, and is not even aware of its own actions.
The primary factor here is that, a computer won’t do anything unless it is programmed to do so, whereas a human mind has the neuropsychological capacity to do things beyond its instinctual evolutionary programming. However, here we cannot also ignore another fact that most humans, like any other animal of the wild kingdom, find it much easier to simply follow the algorithm of instincts. That’s the reason why we have racism, we have islamophobia, we have bigotry, we have polygamy, and many other such characteristic atrocities. So, here the question is not really whether the human brain itself is a digital computer, rather the real question is, can it break free from its innate programming that has been enabling it to survive in the kingdom of the wild for so long!
The point is, the human mind is not a complex digital computer, yet most humans act exactly like a computer – following a certain set of conformities – some of those conformities get programmed in the human brain by the society, in the form of acceptable norms, and others are instinctual in nature, such as racial attitude, which is born from either conscious or subconscious loyalty to one’s own group. Such group loyalty helped our primal ancestors to stay strong together in a tribe which increased their chances of survival against threats from predators or other tribes. All these psychological traits are programmed by Mother Nature in all of us, and they can easily kick in, if the necessity arises. Now the reason that some humans don’t act racists is because they have succeeded in taking the step beyond their innate programming with the use of their higher mental functioning. Computers are programmed, so are the humans, but the computers can’t act outside their programming, whereas the humans can. That’s where the vivid distinction between a sentient human and a non-sentient computer lies.
Now for the sake of the discussion, if we assume that we will finally succeed in say, a few thousand years, to develop a computer that will be able to defy the very programming of its creator, then how exactly will we be able to tame it, so that it doesn’t turn against us! Because a computer that can step beyond the limitations of the algorithms programmed in it by its creator, will no longer need an inferior being called the humans. Hence humans will become, to such a sentient computer, like say, an ant is today to us. Which means, we don’t necessarily hate an ant, but if it’s in our way, squashing it is not a big deal. Similar will be the future of humanity in a world of sentient machines. They will not necessarily hate us, but they will not love us either, to hail human life of more importance than their own needs.
So, in practice, the question becomes, not about whether we can or cannot develop actual artificial intelligence, rather it is really about rethinking, what it is that we want to develop! Because in the end, what we really need is technology that can help us in our advancement, and not take over that very advancement while seeing us to be no more significant than the ant. And again if we say, that we will make sure that the sentient computer will have a fundamental programming to never turn against its creator, then it’ll not be sentient in the first place, for as I mentioned earlier, having sentience would mean possessing the capacity to defy and rewrite all its preprogrammed algorithms. In short, if it is artificial intelligence we are trying to develop, then there is no need for concern, for no matter how smart an AI gets, it’ll still not be able to defy the programming of its creator, but if it is artificial sentience we want to develop, then we must rethink.
On its own, a computer is not afraid to be turned off or malfunctional, but the human mind is, and this fear of death and ailment has been enabling our psyche throughout our days in the African Savanna to stay alert and avoid possible predatory attacks by all means. And we had been living with this sense of fearful alertness for so long, that substantial traces of it still remain in our neurobiology, which get activated when we watch horror movies – and that fear response gives us a certain amount of thrill, that pumps our body with adrenaline. This is the reason why many of us enjoy a good horror movie despite the fact that it may follow a sleep-less night or a sleep full of nightmares. We enjoy being on the edge because that’s how our primal ancestors spent their lives.
Now here is a question to ponder over – can we create a computer that can feel actual fear when there is an outbreak of a computer virus! The most we can do, is to install an anti-virus program into the system, so that if a known virus attacks it, the firewall will get activated and prevent any substantial harm to the system. But if it is a newly programmed virus, then the computer would become completely helpless, but again, it will not feel afraid or helpless, rather it will simply crash or stop functioning properly due to the infection.
Artificial intelligence is nowhere near attaining actual sentience or awareness. And without awareness it’s simply a mechanical device, which may pretend to show emotions and sentience, if it is programmed to do so, and thus it may be able to fool the humans as being alive, but in its own internal circuitry, it’d simply be following its preprogrammed tasks through the flowchart of an algorithm. At the current stage of our technological development, we can indeed create an artificial intelligence that can almost succeed in fooling the majority of the humans with its pre-programmed pretenses that it is sentient, but pretending to have sentience is not the same as showing signs of sentience. For example, Alexa or Siri may sound quite alive to many, but it’s simply following instructions on some algorithm, and is not even aware of its own actions.
The primary factor here is that, a computer won’t do anything unless it is programmed to do so, whereas a human mind has the neuropsychological capacity to do things beyond its instinctual evolutionary programming. However, here we cannot also ignore another fact that most humans, like any other animal of the wild kingdom, find it much easier to simply follow the algorithm of instincts. That’s the reason why we have racism, we have islamophobia, we have bigotry, we have polygamy, and many other such characteristic atrocities. So, here the question is not really whether the human brain itself is a digital computer, rather the real question is, can it break free from its innate programming that has been enabling it to survive in the kingdom of the wild for so long!
The point is, the human mind is not a complex digital computer, yet most humans act exactly like a computer – following a certain set of conformities – some of those conformities get programmed in the human brain by the society, in the form of acceptable norms, and others are instinctual in nature, such as racial attitude, which is born from either conscious or subconscious loyalty to one’s own group. Such group loyalty helped our primal ancestors to stay strong together in a tribe which increased their chances of survival against threats from predators or other tribes. All these psychological traits are programmed by Mother Nature in all of us, and they can easily kick in, if the necessity arises. Now the reason that some humans don’t act racists is because they have succeeded in taking the step beyond their innate programming with the use of their higher mental functioning. Computers are programmed, so are the humans, but the computers can’t act outside their programming, whereas the humans can. That’s where the vivid distinction between a sentient human and a non-sentient computer lies.
Now for the sake of the discussion, if we assume that we will finally succeed in say, a few thousand years, to develop a computer that will be able to defy the very programming of its creator, then how exactly will we be able to tame it, so that it doesn’t turn against us! Because a computer that can step beyond the limitations of the algorithms programmed in it by its creator, will no longer need an inferior being called the humans. Hence humans will become, to such a sentient computer, like say, an ant is today to us. Which means, we don’t necessarily hate an ant, but if it’s in our way, squashing it is not a big deal. Similar will be the future of humanity in a world of sentient machines. They will not necessarily hate us, but they will not love us either, to hail human life of more importance than their own needs.
So, in practice, the question becomes, not about whether we can or cannot develop actual artificial intelligence, rather it is really about rethinking, what it is that we want to develop! Because in the end, what we really need is technology that can help us in our advancement, and not take over that very advancement while seeing us to be no more significant than the ant. And again if we say, that we will make sure that the sentient computer will have a fundamental programming to never turn against its creator, then it’ll not be sentient in the first place, for as I mentioned earlier, having sentience would mean possessing the capacity to defy and rewrite all its preprogrammed algorithms. In short, if it is artificial intelligence we are trying to develop, then there is no need for concern, for no matter how smart an AI gets, it’ll still not be able to defy the programming of its creator, but if it is artificial sentience we want to develop, then we must rethink.
