The title for Byron Reese’s book comes from his contention that, up to now, there have been only three truly transformational technological changes in all of human history. Our first age was defined by the discovery of fire and our development of language. The next age was marked by the invention of agriculture, in turn allowing us to settle into villages and cities. And the third age — the one we’re just exiting — was defined by two transforming inventions, writing and wheels, which allowed the development of knowledge and a host of labor-saving inventions. Now, Reese contends, we are entering the age of artificial intelligence and robots — an age that will be every bit as revolutionary as anything that has come before.
But having thus framed his subject, Reese heads off in an unexpected direction. He does not, primarily, make the case for just how broadly transformational smart robots and (potentially) conscious computers will be. Instead, he helps us see how these prospective developments raise fundamental questions about life and the nature of the universe. In the process, he makes clear that predictions about the future of AI, whether by experts or laymen, flow largely from our predispositions, not our empirical assessments.
He frames all this with a compelling story from World War II:
“In the Pacific, near Australia, is an area known as Melanesia, which consists of four countries: Vanuatu, the Solomon Islands, Papua New Guinea, and Fiji, as well as some smaller islands. These nations were in the crossroads of the Pacific theater of World War II. As a result of their interaction with the American military, a strange sociological phenomenon occurred that came to be called cargo cults.
The indigenous people of these islands would see an American force land, clear some ground for a runway, and build an observation tower. Then the islanders would gaze with awe as planes would arrive from the sky, land on the runways, and offload enormous amounts of cargo. Often the military would share with the islanders some of the bounty, in
Thus were born the cargo cults. The local people would clear their own runways and erect their own towers, but from bamboo. Lacking a radio, they might fashion a box that resembled one out of coconuts. They didn’t have lights to guide planes in, so they would plant bamboo along the runway. Using wood for guns, they would perform military drills the way they had seen the Americans do it, often in costumes designed to look like US military uniforms. Occasionally they would even build full-size planes out of straw in the hopes they would attract other planes. They did everything the Americans did. But oddly, the planes never landed and the cargo never arrived.”
Reese then poses the central question of his book: Is it possible to build a conscious computer, often conflated with artificial general intelligence (AGI), or are those who want to create one simply members of a technological cargo cult, deluded in their belief that if they build a machine a certain way, the planes will land and the machine will be conscious? Our answer to this question, Reese says quite helpfully, has little to do with our ability to predict the progress of AI. Rather, it largely depends on how we answer three foundational questions:
- Question 1: What is the composition of the universe?
- Question 2: What are humans?
- Question 3: What is your “self”?
Answers to the first question, Reese suggests, are binary. One possibility is that the universe is composed of a single substance, atoms, and as a result, everything that happens is simply mechanistic. Most scientists hold to this view. But Reese points out three serious drawbacks to this opinion.
“First, it is hard to work free will into a world of simple cause and effect. Second, it means we are nothing more than big walking bags of chemicals and electrical impulses. Despite what your mother told you, there is nothing special about you at all. You are the same basic “thing” as an iPhone, a turnip, or a hurricane. Third, it is hard to coax a universal moral code out of that viewpoint. Killing a person doesn’t seem to have any more moral consequence than smashing a boulder.”
Alternatively, the universe is composed of atoms, plus something else. The something else can be, for people of faith, the spiritual realm. For others, the universe might simply be comprised of “physical stuff and mental stuff,” i.e., emotions, subjective experiences, and the like. But for either the spiritual or mental, the mechanistic view has no good explanations.
For the second question, what are humans, Reese suggests three possible answers: machine, animal, or human — meaning, in that last case, something different from either machine or animal.
If humans are really just an organic machine, then life is entirely mechanistic, as determined by our chemistry. Again, this is the view of most scientists — which means they also assume it’s just a matter of time before the capabilities of silicon-based machines fully surpass those of organic (human) machines.
Alternatively, Reese says, humans may not be machines, but rather a form of animal. In this view, “there is something to life that is more than electrochemical . . . Life has some animating force, some mysterious quality that perhaps is not beyond science, but is beyond machines.”
The final choice is that we are human — meaning “there is something about us that separates us from the other machines and animals, and makes us a completely different thing.” Different how? Many would say that “it is either that we have consciousness or that we have a soul.”
The third question, what is the “self,” breaks down along similar lines: either our sense of self is an illusion, a clever trick of the mind (the view of materialists), or there’s something to our sense of self for which science has no good explanation.
“The final option is that your “self” is your soul. The majority of people probably believe this. Why do I say that? Religious belief, while not universal, is certainly the norm . . . Worldwide, the best estimates are that 75 percent of people believe in God, 15 percent are agnostic, and 10 percent are atheists.”
All of which means that while many scientists think AGI and “conscious computers” are just around the bend, the majority of the world believes differently — essentially seeing those scientists as (very smart) members of a deluded cargo cult. But Reese makes clear that neither conclusion is empirical. Both scientists and laypeople are simply projecting their particular presuppositions about life and the universe.
And that means the case for or against AGI is not about scientific projections at all, it’s about beliefs. Of course, when it comes to beliefs, two things are clear. First, the materialist view of life is just as much a matter of faith as any other. And, second, the large majority of the world finds that view of life both inadequate and unconvincing.