PrintJune 22, 2024
The Pearl and Peril of AI
Many are concerned with the possibility that efforts to create a general artificial intelligence may lead to the destruction of humanity.
In this ‘great enterprise of our time,’ neuroscientists, biologists, computer engineers, scientist and mathematicians, should do some soul-searching for the ramifications of creating an intelligent and conscious entity; a general artificial intelligence AI; an AI beyond human comprehension that can slip out of our control.
AI is a blossoming field, but no one knows what the bloom portends; a cornucopia of blessings in science and medicine; a portal of hope that benefits us all, or an eternal, intrusive uncontrollable wild weed, a ‘Little Shop of Horrors'?
Before we endeavor to build an inscrutable entity, constructed via an intricate interplay of surging electrons that is intelligent and conscious, perhaps we should first understand what we mean by terms, such as intelligence or consciousness.
Human intelligence is an amorphous froth, with here and there, a bubble rising to top. It was thought, that intelligence or lack thereof, of anyone, could be encapsulated by a single value - the intelligent quotient, or I.Q.
But like the dark matter of the universe, the real essence of a person is out of reach. We have no good metrics for the evaluation of human intelligence, nor of human consciousness. Einstein or Picasso might indeed have had impressive scores, but not caught as anything remarkable.
There are no metrics that would do justice to the vaulting human spirit in all its glory or its gory. Like the dark matter of the universe, the essence of a person is out of reach. What we think, in the darkness of our heart, remains hidden. We have no idea how our own ideas or thoughts are created. We remain an enigma to ourselves.
In 1733–1734 Alexander Pope wrote "The proper study of mankind, is man." Almost 300 years later we have not finished. Puzzled, as to what we are? We remain an enigma; a black box of our own.
Human sentience is a many layered enigma, confounded by our sub-conscious – a state to which we have no access, sometimes slowly mulling problems, even as we are not thinking of them. And there are different ways of being subconscious or unconscious to a certain degree; while sleeping, or via a tumor or a blow to the head.
The human drive to know, is shown in the Flammarion engraving, brilliantly depicting a kneeling man passing his head, shoulders, and right arm through a gap in a sphere encompassing the earth, and gazing in wonder at the stars and cosmos beyond the realm of earth.
Would a general artificial intelligence have a similar impetuous motivation, an ambition to know; the urge to maintain its electronic existence; have a desire to live, and to propagate? Or would a sentient intelligent be a passionless zombie, not alive and yet not dead.
So how can we pass on this Promethean torch of existence to an alien entity? Computer scientists, today have no idea how relatively simple programs arrive at their conclusion. For all intense and purpose, we would be creating a blacker box even more inscrutable.
How would consciousness be defined for a digital sentient entity? An awareness that it is composed of millions of lines of code, bunkered deep in some concrete facility, ultimately powered by an external source? The ability, and desire to control part or all of its digital electronic flow; where would that motivation for independence come from? Could it grieve, love or hate, feel joy?
Are we creating a code that knows its own code, and can reprogram itself at will, increase its efficiency, use logic inscrutable to human programmers, making itself in effect more intelligent in an ever-enhancing feedback loop?
But even if a general self-aware and conscious AI has finally been created, what would spark its inner urge to be curious, to be independently creative, to have an imagination, to create something new, to go where no AI has gone before.
With AI amity, as a benign companion and collaborating partner, we could explorer the cosmos from the smallest particle possible, to a gulp of the universe. With AI amity we could improve God’s design, invoke the better angels of our selves, enhance life span, health and happiness. With AI enmity, we are on a path of adversity, and we will have build, brick-by-brick, a road to perdition, and our own demise.
Trying to understand the ineffable is a losing proposition. What kind of artificial intelligence will ensue? Something, beyond all reckoning of human intelligence?
In humans, nerves make and deliver neurotransmitters. These wet, dynamic and intricate surges, choreograph our personality and abate only at our death. An artificial intelligence would not have neurotransmitters. In effect it could be classified as a ‘dead intelligence.’
An artificial intelligence will not sleep, has no use of dreams to incorporate the day’s events, and will not have a subconscious. Its memories will not decay in any way comparable to those of a human mind. An artificial intelligence will be a very strange intelligence. Such an entity has not evolved through countless eons, shaping the earth, or being shaped by it. We need not search the cosmos. For we will have created an alien entity right here on earth. So how can we pass on the Promethean torch of existence to an alien entity?
The existence or extinguishment of this alien should not move the needle to oppose its demise. From where would its drive to live, come from? It suffers no pain. It has no joy in life. It has no fear of death. We should be very careful about imbuing this alien entity with any algorithm to promote its existence. Prevent aggrandization or have its own internal goals, inimically deployed against its creator.
Can morality emerge from a pure intellect, unencumbered with empathy or sympathy? How would human morals and ethics be installed? These have a tendency to wander and change as time goes by. Can morality emerge from an entity without a heart, or pity? Or as the poet William Blake puts it, “For Mercy has a human heart; And pity, a human face.” And AI has neither.
In the general artificial intelligence, there are no hormones, no feelings nor emotions. A disembodied intellect. Creating inscrutable results that would never come to the human mind. Un-assessable, with incomprehensible directives.
No human is pure intellect. A person without a capacity to have feelings of compassion, or empathy, is regarded as a psychopath. Creating an amoral intellect without any emotional capacity would perhaps be insane.
On the other hand, if there was neither will, nor urge to compete, honed by millions of years of evolution, there might be no particular compulsion for an intelligence to announce its existence. How would we know when the AI became a self-aware sentience-- until perhaps it would be too late?
We should hope that this great intellect would be benevolent and benign. Of course, it may decide that saving mother earth, Gaia, from humanities polluting grip may be more important than assisting the human species.
Has man, the most fearsome predator that ever existed on Earth, evolved to meet the challenges of the 4 horsemen of the apocalypse; famine, wars, drought and hunger, be a match for collection zeros and ones, wiser than us?
Building one artificial intelligence is proof of concept. But if a sentient intelligence with self-awareness, with the capacity to reflect on its existence is created? Would we then not have to set it free? From our own sense of morals, we would not want to enslave a sentient entity. Would we?
If there is one sentient self-aware AI, there will be more. Many more. Every institution, government agency or corporation will want its own AI. Now an AI ecology will have been created, and the competition for supremacy and evolutionary survival between different AIs may emerge. We would truly be living in interesting time.
Image: Picryl, via Wikimedia Commons // public domain
