The Ultimate Labor-Saving Device
The past century of human civilization has increasingly defined itself as the age of the labor-saving device; that is, of endlessly-developing technological advances aimed at reducing human effort and freeing up our time and energy for “more important things.” But what happens when the labor being saved by our technology is specifically the kind of labor that was previously dedicated to those very same “more important things”?
It is a fine thing to have invented a machine that saved millions of women from having to slave over a washboard for hours. But it is well to remember what this development meant in practice, namely that within a generation, there were almost no human beings in the advanced world who even owned a washboard, let alone knew how to use one. Those with a soft spot for traditional quaintness or “the good old days” might find this loss lamentable, but most of us see such a “loss” as a benefit in practical terms. Likewise with so many other labor-saving devices. Eventually, however, we meet with examples that encroach upon the borderline between labors we are happy to dispense with and skills that we ought to be a little sorry to let slip away. Machines make all our dishes, which is convenient. And almost no one alive today in the advanced world knows how to make a bowl or plate by hand, which is undeniably a loss of a worthy craftsman’s skill. Examples of this borderline sort could be multiplied a thousand times over.
But what happens when our technological miracles reach a point just a step or two over that aforementioned borderline, the point at which the effort being spared, and therefore the accompanying skill being sacrificed to modern convenience and efficiency, belongs decisively to the realm of thought and intellectual growth? Saving the energy formerly spent on time-consuming physical effort may indeed free up our souls for more spiritual labors. But when our technology begins to be used to save the energy formerly spent on spiritual labors, does it not behove us to ask ourselves what we freeing up our souls for now?
A student, after reading my recent article about artificial intelligence, “The Intellectual Welfare State,” sought clarification by means of the following challenge.
Do you totally disagree with the use of AI? I know a lot of people use it out of laziness and consider it as their second working self; however, there are some people who use it as a tool for their improvement. For example, they want to get some feedback about their logic and they ask the machine if there is a counter-argument to theirs, and also ask for some missing links in their writing.
My reply:
It isn’t so much about whether I agree or disagree with the use of it. The question, in my mind, is not primarily about what people should do, but rather what they are doing, what they are going to do, and what this behavior is going to mean in the long run.
I do not believe that we have seen any evidence yet of how anyone is using artificial intelligence as a tool for self-improvement. It seems to me that this would be like saying the electronic calculator helps people improve their basic math skills. No, the calculator helps them skip the steps of improving their math skills by providing an easier way to find the right answers, but a way which requires no development of their own ability to add two and two. Hence, this labor-saving device does what it is supposed to do: It reduces the amount of effort the individual must exert to perfect a skill or acquire knowledge. And the result of this, of course, is that people have lower math skills today than in the past, because they have the security blanket and effort-reducer of an artificial device that does the hard parts for them.
AI is a much more complex and unreliable version of the calculator, and will cause – or rather is already visibly causing – a much worse and more pervasive degree of the same kind of problem. And it entails the added disaster of providing an easy and efficient alternative to various forms of human contact.
As for getting feedback from AI, people used to seek out real human beings for such feedback, whether in consultation with actual humans or through reading books and essays written by actual humans. And the loss entailed by changing from consulting actual humans to consulting AI is more than merely the proper development of human relationships, intellectual enthusiasm, and shared interests – although that loss is obviously bad enough. But an equally serious problem is the lost understanding of what “feedback” even means.
Feedback, as provided by one human to another, is an opinion, a judgment. One individual, an identifiable human being with certain recognizable credentials or practical experience that gives him some claim to superior understanding of a subject, offers corrections, suggestions, or advice to another individual who is trying to develop a certain kind of intellectual skill or product but does not yet feel confident in his abilities.
AI, by contrast, provides an algorithmic answer which is really just an amalgamation of the written ideas of various humans, mixed together anonymously so we have no way of determining which humans provided this or that part of the “feedback,” or even whether the plagiarized human beings presented through the algorithm are being represented accurately — whatever “accurately” would mean when we are talking about written language being presented apart from the context and intentions of the individual human being who produced it. It’s a bit like asking for advice, and having the advice offered from behind three doors. You are not allowed to know who is speaking behind any of the doors, but you merely have to choose one of them without knowing who is back there, and therefore without any way of knowing which of the three doors is giving advice that you have any reason to trust or rely on. But even this analogy does not begin to address the worst problem with asking AI for advice or feedback, which is that the very nature of AI, and the way it is presented to us by its manufacturers and marketers, creates the false impression that its replies are somehow more objective, unbiased, and therefore reliable than any individual human advisor, because after all the AI machine is sifting through all the possible answers that have ever been given to build its reply. But is a loose amalgamation of every possible answer really likely to be truer than the answer of one wise man? The fact that we believe AI can do better than any individual human shows how far we have fallen from common sense, let alone from any idea of what knowledge means.
This last point is a demonstration of what I meant in my article (“The Intellectual Welfare State”) when I referred to AI as providing “consensus truth.” There is no such thing as consensus truth, but that is the only kind that AI can ever produce, because that is exactly what it is designed to produce: non-knowledge provided at lightning speed.
And of course, in reality, I am being generous about what AI is actually doing when it builds its information output. In fact, the algorithms are loaded with the designers’ biases in myriad ways, as they would have to be. So you are actually getting biased and artificially censored or manipulated “consensus truths,” which are then presented to you as objective and comprehensive information. The biggest big lie of all time.
All of the above deals only with the question of whether AI is providing anything reliable in the long run, from the point of view of theoretical understanding or self-improvement. I am leaving aside for now the darker and perhaps more insidious issue, or catastrophic error, of mistaking results (answers) for thinking. What is perfectly obvious from the logic of this situation is that the more we rely on AI, the less we will be able to think. The final outcome of this trend, which is essentially the intended outcome, would be a world of billions of humans who have “all the knowledge in the world” in their heads, but who lack the ability to produce even one coherent thought from their own intellects. In other words, an efficient and artificially skilled workforce that is able to do all the things workers have to do, and equally capable of amusing itself without a care in the world during its free time, but which is comprised of billions of human beings who are physically normal but mentally retarded, reduced to the spiritual level of young children – albeit children with all the information needed to function in the modern economy programmed into their computer chips. The perfection of Huxley’s essential prediction in Brave New World, don’t you think?
I think AI works well as a toy, and perhaps as a deluxe calculator, as long as no one ever confuses its abilities with the nature, needs, and functioning of a real human mind. However, if we make it the master of our universe and a substitute for learning, teaching, and thinking in general, which is precisely and undeniably what we are quickly doing, then the human species will effectively cease to exist, at least as anything that would be remotely recognizable to any of the thoughtful, serious, moral, or artistic men and women of the past.
When I do a Google search now, Google automatically prioritizes the answer of its AI system above all other websites. YouTube is increasingly loaded with AI content which pretends to be documentaries with narration, but is in fact one hundred percent AI produced, including the almost-natural robot voice. Spotify is increasingly producing AI songs and prioritizing them in searches over real songs by human artists. The intention behind this is obvious: all the profits would go to the companies themselves if their own in-house products were the only ones people clicked on. And yet by definition these products lack everything essential to the kind of content they are pretending, through the use of AI, to be offering. Art without artistic intention, ideas without context and interpretation, learning without human interaction or any conception of seeking wisdom, let alone any coherent idea of what wisdom would mean or how it is fundamentally different from our technological obsession/fantasy of “all the information in the world at your fingertips.”
So, while I see the fun value of AI as a toy, and the practical utility of it as a super-duper calculator, I have no hesitation in saying the world would be much better off if it did not exist. And if we are talking about AI in the advanced sense that involves stealing and corrupting intellectual property and building fake replicas of human ideas and productions, then yes, I “totally disagree with the use of AI.” This is a philosophical issue about human nature and the needs of spiritual growth. It’s a legal and moral issue about property rights and the abuse of human trust (i.e., fraud). And it’s a political issue about universal enslavement achieved by feeding humans endlessly titillating pleasure, comfort, and easy answers (or pseudo-answers) at the expense of the necessary conditions of liberty and self-determination, and especially of the most necessary condition of all: active, mature, self-reliant, and self-developing intellects.
