Why it's worthwhile to care about AI, Pt. 1
Even (and in fact especially) if you aren't a crazy materialist transhumanist!
Despite the fact that I am a lover of things mathematical and scientific, for a long time I didn’t care about AI. I really didn’t. In fact every time I heard the term, my eyes would nearly glaze over as I began picturing weird looking white humanoid robots, far fetched science fiction stories about a new master race, and bald nerdy looking computer scientists who hadn’t been out of their basements for weeks talking in excited tones about futures in which the entire world’s population was one giant cyborg mind, ruled by some sort of pure, “unbiased” machine of our own creation. The whole thing simply had no appeal to my imagination.
Fundamentally, I think the sticking point was this: all of those most excited about AI seemed to be excited based on a false premise, namely a materialist view of mind (and reality) which views mind and intelligence as an emergent phenomenon, arising when some threshold of logical complexity is reached in a purely deterministic machine. According to this view, you and I are both just machines, albeit complex and well developed ones. Taken to its full logical extreme, this view must posit that any ideas we might hold of spirituality, eternal life, free will and moral heroism, love, perhaps even consciousness and ideas themselves — all these are mere illusions, machinations of a system predetermined by its initial conditions, “developments” in a ruthless zero sum game of survival of the fittest as little gears and circuits strive to hold on to that thing they have found expedient to call “life”.
The funny thing is, though, that even the most extreme materialists still manage to find some (sometimes quite high) level of joy and wonder in life. So there’s a sort of sleight of hand at play: reduce all life to the level of a machine in a way that seems to fly right in the face of our own experience of our lives and freedoms, while simultaneously holding it elevated at the level of desirability and goodness with which we experience it. And the biggest winner following this little trick? So called artificial intelligence! If everything we know of our own lives is in fact mechanical, then it can at least theoretically be copied, reproduced in the workings of machines of our own making. And if, simultaneously, life and mind continue to have the value which we place on them under the assumption that they are not mere mechanically determined processes, then, voila! it follows that the pursuit of artificial intelligence is the pursuit of a machine capable of doing and encapsulating all the most wonderful things we can be or experience.
People are motivated in all kinds of different ways. The strongest, best, and most sustainable way, I think, is motivation in the true pursuit of a lasting and deep good. This motivation is good both because it is lasting and not temporary, and also because what it leads towards is worthwhile and not warped or evil. But illusion can also, at least temporarily, be deeply motivating. Think of the addicted gambler who faithfully buys a lottery ticket time after time, convinced that the very next one is going to be that lucky one which unlocks an entirely new life for him. Or, perhaps a closer analogy for the biggest AI enthusiasts might be the alchemists or astrologers of old. The interesting point with the alchemists, for example, is that while they were incredibly motivated by false hopes, like the hope of turning base metals into gold, they also wound up doing a lot of fruitful work, laying down the initial foundations of chemistry. Similarly, the early astrologers attributed a causality to the motions of the stars which did not in fact exist, but they did also manage to get a pretty good understanding of the motions of the stars, beginning what we now call astronomy. And so in the same way, in my opinion, are the current AI enthusiasts led by a false hope of recreating in a machine the full goodness of human life, but nonetheless led headlong into the pursuit of what could, in fact, be a truly incredible and positive innovation.
What’s important, though, is that at some point the immature pursuit of the illusion hands over the fruits of its labors to a more mature pursuit of what is good and worthwhile, based on true reality. And this is the juncture we are rapidly reaching with AI. Needless to say, the motivations of AI researchers have always been mixed, and some very level headed people have been involved throughout. But a tremendous amount of energy continues to be focused around materialist premises. From promises of complete replacement of the entire human workforce to fears of rogue terminator robots, the loudest voices are still almost all thinking of AI as a force which is the same in kind as human beings.
And this is where you, my sane and wise reader, come into the story. What we need now with regard to AI are people — smart, wise, and many of them, who understand AI, appreciate it, and will help shape and build it while maintaining a clear sighted grasp of its proper role and limitations. Deep, soulful, courageous men and women, possessing of something close to the fullness of humanity in themselves, people with a heart as well as a mind, who are able to see the power of AI and to situate it properly as a tool to further genuine human thriving.
Doing so will not be simple or easy. AI is in fact almost insanely powerful, capable of truly incredible and surprising things. It likely will be the most powerful artifact created thus far in human history. And it genuinely has proven itself capable of doing things which were hitherto thought strictly the province of human minds, things like generating art, or winning chess games, or speaking like a human, or offering practical advice. The conclusion one ought to draw from these things, however, is not that humans were in fact mere machines all along, but rather that what makes us human is deeper, subtler, and even more valuable than hitherto thought.
This last fact, that AI casts into high relief what it truly is to be human, is well worth exploring, so much so that it will have to be its own post at some point in the future. But in the meantime, if we are to really understand how we humans ought to use and shape AI, and what really makes it worth caring about in its own right as well as in its contrast to the even more worthwhile fully human life, we are going to have to delve a little deeper into what exactly this thing is. And that too deserves a post in itself, and so will have to be part two of this series. However, a few things we can say now by way of summary of this post and anticipation of the next:
AI will not be human. It won’t even be in the same category as human beings. Human beings are living, loving beings with both body and spirit, capable of comprehension, wisdom, free choice, love, virtue and (though you need not agree with me about this to get the point of this post) eternal life and divine worship.
Computers, and computer programs, including so called “artificial intelligence,” are machines, made by humans. Like all machines, they are inanimate, physically determined, non-conscious, amoral, controllable but unpunishable, intrinsically unaccountable and a-responsible, finite, and subject to degradation under the 2nd law of thermodynamics — which is to say, neither possessing nor capable of any of the above things.
The idea of AI appeals the most (naturally) to people who overrate it, which happens in this case to be materialists who can’t see the above two points. As a result, materialists are significantly overrated amongst the ranks of AI pioneers, which can lead others to be turned off by the entire project.
Despite this, and as we will see in more detail in part two, AI’s are (or at the very least soon will be) the most perfect machines thus far created by man. As such, they are also the most powerful artifacts yet known. AI, while not capable of itself having a will, certainly is going to be capable (as all artifacts are, though none to the degree that AI will be) not only of carrying out the wills of real people, but also of having causal effects on the wills of people. Moreover, being extremely new and powerful, AI will have (and already has had) unforeseen effects of the first order. One might think of architecting a city as a close analogy here. An ugly city does less than no good for the hearts of those who come into contact with it. A good and beautiful city, on the other hand, is a truly wonderful thing, ennobling and uplifting the spirit in a way that is hard to overstate! And in either case the city will likely have statues, likenesses of people, both indicative and formative of the aspirations of the city’s inhabitants.
In a way that is partly like giant digital statues, and partly like the entire city itself, this incredibly powerful machine which goes by the name of AI will indeed bear the image and likeness of its creators, operating vaguely according to their wills, and furthering their view of the world, far more than it does the wills and views of its proximate operators, and certainly more than it does the wills and views of the average citizen. This is true especially now, as we have almost no mechanisms of any kind of deep shared governance of this technology, nor even the common level of understanding of it which would be required to truly democratize it.
In closing:
While not human, AI will be made in the image of certain aspects of humanity, and it will be a nearly perfect machine, which, as we will see in the next post, is a lever of power and attention by which the power and caring attention of those who control it will be vastly, vastly multiplied. Which image, exactly, shall we choose to recreate, and whose care and attention shall we multiply? The stakes could hardly be higher. The choice, ultimately, is in our hands. But the caring attention must be paid. The adventure of building this new tool and context for human beings, much like the city building and frontier settling of past ages, is full of possibilities for both good and ill, and beckons to the bold.