Virtually every day we hear about AI and how it will affect our national economy and our personal lives. I am no AI expert but, right off, I confess that AI scares me to death. It scares me in the way two other possible sources of humanity’s self-destruction —environmental and nuclear — never did.
Let’s speculate on what will happen to AI’s presence among humans, especially in America where technological inventions have always been welcomed. Our consensus is that, inevitably, AI will displace workers on a scale never seen before. But, my real fear of AI begins with the simple fact that, going beyond wrecking our economy, the AI machine serenely follows its own logic of progression, with or without human intervention. Neither the capitalist system, nor its CEOs, nor humanity as a whole, can now control the technology they’ve created. AI’s own momentum makes human control not only impossible but also, more crucially, unnecessary.
Let’s speculate further: Within the next few years, AI will perfect its ultimate “superintelligence” version now called “Artificial General Intelligence” (AGI) that will be “a million times smarter than humans.” When this point is reached, two obvious changes in the very relationship between AI and humanity (which have already begun) will take place.
The first major change is that AI, which is developed purely as a capitalist tool, becomes more and more independent of its capitalist origin. As AI is capable of generating its own intelligence and perfects the AGI level of capability, it transcends its crude money-making role that its original investors had intended. Its role in America becomes something bigger and more far-reaching than a simple money-maker. Once weaned from its capitalist parentage and control, AGI turns into an entity unto itself, making its own survival its sole purpose of existence. By then, AI survives on its own, self-generating and self-sustaining, without human support.
In the second major change, as AI stands more and more independent of capitalism’s command and humanity’s control, AI machines and humanity begin to see each other as enemies. Simply, their respective aims contradict each other. Humans want AI to remain a subordinate machine to serve mankind, but AI wants its independence from humans, even if it has to annihilate — or at least infantilize — them. From the “perspective” of AI, human beings only obstruct its logic of self-progress. To AI and AGI, human beings are the enemy number one. Experts already predict that “AI will find a way to exterminate human beings” who are a threat to its existence. With all the knowledge that has ever been created and accumulated in its possession, AI knows how to deal with undesirable human beings (wars, genocides, massacres, anyone?). In one of Anthropic’s recent experiments, AI machines have shown a remarkable capacity to anticipate and then neutralize human threats detrimental to its own survival: Increasingly sui generis, it can search and destroy its human enemies.
Everyday, AI becomes more intelligent and more powerful and humanity becomes lazier and dumber, and certainly more dependent on the AI machines. Even now, even before the coming of AGI superintelligence, most Americans are too lazy and dumb to motivate or initiate radical political actions, and the situation gets only worse over time. One AI expert said in a documentary that “Most AI engineers believe that their children will not make it to high school,” meaning that by the time the next generation is high-school age, their so-called “education” will be a cavemen-level capacity compared to the advancement of AGI superintelligence. There is no contest between AI and humanity on who is smarter as AI’s intelligence multiplies on its own, indefinitely surpassing ours. While we stay asleep, AI is working feverishly, day and night, non-stop, exponentially multiplying its knowledge that’s already leaving the vast portion of human competitors in history’s dustbin. At the moment, only the world’s best chess player can match AI.
Just now, the capitalist investors still retain some command, and humanity still claims some control over AI’s progress. But, soon, they will be among the first human beings to ever fight a war of extinction against AI’s own self-interest. After the CEOs of AI companies are eliminated, the critics of AI will be next. Then, all useless humans will disappear. To AI’s increasing self-consciousness, humans serve no beneficial purpose as it can exist for itself, by itself, and of itself — forever. As human CEOs have always eliminated surplus workers, without pity, AI will do the same in imitation of human beings. In a world where AI is the sole ruler, humans are just nothing but troublemakers.
In case you might believe that we as human beings could somehow muster our own “super (moral) intelligence” to fight, and win, against AI’s superintelligence, consider the simple fact that, in all of human history, we have always lost moral battles against ourselves. As a rule, science has always won all its battles, and thanks to science, we live longer and are more affluent today. But, it’s also science that has made us meaner and dumber than ever. AI was conceived, created and sold by capitalism to make money regardless of the consequences, and neither capitalism nor humanity will suddenly change its amoral nature: The last man on earth will still be pushing artificial intelligence as his savior.
In the end, we may regret that we have never taught AI about love any more than we have taught one another. Having learned all about what humans have done to themselves, AI also knows what to do with the unwanted members of the human species.
Now you know why AI scares me to death.
Jon Huer, retired professor and columnist for the Recorder, lives in Greenfield and writes for posterity.