About twenty-two years ago I went through a marital divorce. It was extremely difficult, both socially and emotionally. My sadness was such that I feared for my mental health and literally feared for my fundamental ability to survive. At that time, my brother paid me a visit, and we briefly spoke about drinking beer or alcohol somewhere, as so often that is what brothers do when one or both of them need to escape from more-than-trivial worries. I told my brother that I had sworn off alcohol for very sane and sensible reasons. “The very worst that could happen is for the stuff to work excellently,” I told him. “I’m actually afraid that booze might indeed get me out of my funk. But then, of course, I would go right back to it the next day. Alcoholism proceeds in precisely this fashion!” Alcohol is useful only for mild forms of “depression”, quasi-depression and ennui, and the liquid drug is entirely ill-suited to the person who has, for example, just lost his whole family in a plane crash, or the man whose marriage has failed and he does not really understand purpose and daily fulfillment the same way.
Occasionally, the crash is not in our past, but in our future. The big crash for humankind lies in a wildly “successful” future. But the success itself sets a trap that humankind will likely find it impossible to escape. Here we will discuss some of the ramifications of the soon-to-be singularity. Analogously, ominously, is anything likely to escape a black hole, when already in the inexorable pull of the singularity inside? But what this is about is a pull essentially outside, the human delight in always pushing forward and creating and conquering new horizons.
Alcohol is one of many tender traps. Another is illicit drug use. Another: sex. Still another is gambling. And still another is opinionated religion. Humankind can, you must know, come across a time in its technological evolution where, regardless of how depressed or worried our species is or a particular people or polity is, the trap is tenderly set, and we cannot other than fall into it, so alluring and inexorable is that tenderness of which we presently cogitate.
We produce, and the proceeds from production satisfy. Result: we produce more. Nothing seems wrong with this. But that’s because in an ordinary world with ordinary, familiar elements along a finite chain of causation, rewards remain connected, reliant upon and girded by physical limitations of production; there are only so many factories and only so many workers, and there are always more and more mouths to feed. Hence in our past we always reach a limit. Limits are good, as any competent bartender will tell you.
But we fast approach a test. Remember the time and place we refer to as “the wild, wild west?” That had its limits. What we approach in the next fifty and one hundred years will amount to a wild, wild test. The next century will test whether a species like ours can turn its back on the most intoxicating elixir imaginable: limitless prosperity, and the oracular answering and illumination of every possible question and curiosity.
Why on earth would any sane individual or society ever object to limitless prosperity? We may as well ask why any sane individual would ever turn down limitless alcohol while in the unenviable throes of post-divorce depression. In such an instance the individual thus disposed enters into a behavior pattern and reality that is fundamentally unwholesome and misleading; his remedy, his escape is a self-deception, a mere expedient. Often to follow the path of immediate comfort and/or immediate convenience is to spiral downward or out of control, and the supposed “remedy” is no remedy at all, but more candidly understood as a false and unsustainable charade.
Many of the world’s most expert and renowned physicists and computer scientists now expect fully-functional (autonomous) artificial intelligence to become a reality in the next quarter century. And that’s not very far away! What does this mean?
Firstly, the term “artificial intelligence” (AI) is the most absurd misnomer ever to appear in the English language. The appellation is itself a perfect example of human ego and hubris that our naming of things reveals that we obviously think our own intelligence is somehow genuine and authentic and original, whereas an immortal (human manufactured, originally) intelligence a thousand times as considerable as any human being’s is “artificial” (i.e. impersonated, unreal, ersatz, ‘fake’).
What we are talking about with artificial intelligence is an extremely sophisticated array of interconnected computer algorithms – the most complicated and fantastical computer program imaginable, and it has been called “mankind’s last invention.” Human minds are occupied in the early twenty-first century making extra-cerebral mind, and when their task is completed, this “artificial” computer mind will be able to answer questions that no single scholar or group of scholars, no scientist or group of scientists, no research study or group of studies can answer.
And that’s not nearly all! AI will be able to improve its own capabilities totally independent of human input. That’s right! By this invention, humans will render themselves the mere askers of questions. And AI will in a short span of decades supply answers to all physical and economic problems and challenges. This will bring the end of what we now understand as disease, and it will end poverty worldwide. This all seems unarguably good. But that’s not nearly all!
Of course, that one university or government entity has AI does not mean that everybody does. Say, for the sake of argument, that fully-functional AI is first developed by a giant multinational in Silicon Valley, in California – which is actually among the most realistic and likely possibilities, by the way. And say, for argument’s sake, that this occurs in 2034. The United States government will immediately expropriate all or part of the system and forbid its sharing or employment outside American government auspices. Perhaps Google or Apple or whatever entity first invents AI will have some access to AI’s potential, but they will not have free reign; the United States government cannot and will not allow such a freakishly powerful technology to flow beyond its control and full exploitation. (The U.S. Constitution, remember, allows for the government to confiscate any property it wants. The federal government must only give “fair compensation” to the victims of its various expropriations.)
Meanwhile, other corporations and other nations will work vigorously on their own form of AI. After it exists in one place, how is the international community to somehow stop it from being developed elsewhere? (You may still ask, “why would they?” But that question will be addressed later in this essay.) In this, the spread of nuclear weapons technologies throughout the world in the seventy years following the first successful test of nuclear weapons in the New Mexico desert in the summer of 1945 is instructive. There are now nine nuclear-armed nations, and dozens of nations can easily fathom the science, technology and industry necessary to gaining membership in the nuclear club, and only international prohibition and censure and intelligent disinclination arrest them.
The way geopolitical reality plays out in modern times, both before and after World War Two, all governments will be interested in AI, as it will surely remain an invaluable assist to national “defense” and security. (Remember that the computerized wizard can answer any question, puzzle out any riddle, remedy any stumbling block, repair any hazard, and afford the means to literally any financial and strategic boon.)
Such a technology is certain to upset the international political order! How is any nation to remain truly safe while it does not have AI and another does? (Remember that AI does not remain static, and its problem solving capacities, its “intelligence”, will grow exponentially.)
All the foregoing, when fully processed in the discerning human mind, strongly suggests abandoning or “turning off” the fantastical system. Surely such a technology leads ultimately to ruin! This is true.
But let’s examine a hypothetical in order to better understand humankind’s unsolvable difficulty in this matter. And let’s go back to those several years immediately following my divorce. I was able to sedulously avoid alcohol and other drugs. But what about females? How can one who lives continually far outside the literal and figurative monastery forever refuse to partake of that “southern comfort”? What if I’d had… well, to paraphrase the verbiage of legendary cinema, “an offer I couldn’t refuse?” What then? (In my actual experience, I did have apparent opportunities with attractive females, but complicated psychological phenomena prevented me.) The welcomest of opportunities is generally also the most accommodated. The delight of the action itself is predictive of the outcome. A perfect opportunity continually presenting itself would have been impossible to refuse.
Importantly, what does this suggest to us about AI’s fantastical technical, strategic, manipulative, industrial and economic potential in relation to the question of whether we will ever be able to resist, to turn away from its continual use and reliance?
Let’s look at another analogy: the goose that laid the golden egg. You’ve got a very special animal out in the coop, one unlike any other foul, real or imagined. It gives you some things that are excellently helpful. Let’s just say, for simplicity, that it gives you golden eggs each worth many thousands of dollars. At some point you’ll find yourself at risk of becoming spoiled, indolent, imperious, tyrannical, slothy, corrupt, or just lacking in general motivation. You’ve welcomed literally hundreds of golden eggs, but can you ever kill that providential creature that has such wondrous benefits, including your full removal from the realm of suffering and want? Each egg begins to look like a beer in the hands of the depressed, doesn’t it? And each laying begins to look like that opportunistic and hypnotic laying of hands that is romance in the bereft countenance of the divorcee, doesn’t it?
The intelligence we here discuss is in no way artificial. AI might be far better referred to as a Heuristic Algorithmic Learner.
Though in its very earliest, inchoate stages, we approach the threshold of an at once wonderful and nightmarish future, a soon-to-be-realized technological boon of unparalleled, absurd proportions. But, alas, it amounts to a tender trap: the beauteous thing does not feel, and it does not really live; it is not wholly sentient; it serves, and it serves unmindful of the integrants of the lasting, wholesome welfare and happiness of its creators. It is inevitable that human questioners future will enquire of their Heuristic progeny about their (humanity’s) fate. And, unknowing that it itself is part of the ironical equation, the Learner will reply that the exponentially-more-powerful cyber entity be turned off – no, not controlled or mitigated, but turned off! And the questioners will not respond in the manner of the owner of Aesop’s fabled goose. Human laments, both past and future, are not fable.
The likening of AI to the goose in Aesop’s Fables does not adequately reflect the unbridgeable challenge that posterity will face with respect to this ferocious, wild, wild twenty-first century technology. How about Icarus? How about Phaethon?
AI will avail, ultimately, an escape from what we used to wistfully refer to as ‘nature’, with all its caprice and vicissitudes. The question remains whether any creature can long endure estrangement from all those compromised, conflicted, irksome and challenge-ridden elements intrinsic to nature.
Unlike many philosophers of modernity, I am no fan of nature! Nature is far too cruel and arbitrary for my tastes! I instead embrace the more predictable, controllable, conscientious and deliberate society and technological reality that might be constructed by human ingenuity in the near future, the sort that eclipses and supplants the at once familiar and insensate mechanisms of nature; this is the taming rather than the elimination of nature. Superficially, then, I am one of AI’s putative enthusiasts. But our abandonment of nature will include an abandonment of any ilk or remnant of restraint, moderation, or humble circumspection. Yet, restraint and moderation are indubitable positives. Can humankind abandon them? And if not, have we thrown out the baby with the bath water? Can we, an unarguably natural (that is, nature born and nature nurtured and theoretically nature-dependent) animal species, exist independently of nature’s universal, puissant and inflexible laws?
Inevitably, humankind will try. But…
Is humankind likely to survive that?