Who's Afraid of Big Bad AI?

My wife and I get our groceries delivered by a big online company that services a large part of the New York metropolitan area. The other evening I attempted to add a few items to an order I had placed earlier but received a notice in return that the company's software was malfunctioning. I decided to avail myself of the "chat" option offered, and after a few minutes got a message from "Kirsten" saying she was ready to help me. I explained I had been attempting to check out after altering my order and was unable to so do. She replied that she could add the new items to the order for me. That wouldn't work, I told her, because I couldn't remember all the products I had added.

We went back and forth like this until she finally apologized for the mishap and said the tech people were working on it. I returned to the site and tried again to check out. This time it worked. I went back to the chat box, told Kirsten what had happened, and thanked her. She said she was glad I had succeeded. I thanked her again and logged off.

Was there really a person named Kirsten or any other human being on the other end of that chat, or had I been talking to a robot? I think it likely Kirsten is a real human being, but from what I'm hearing, it won't be long before it will be impossible to tell whether she is real or a robot.

Back in the 1960s someone devised a computer program that simulated the personality of a psychotherapist. The idea was to see if a human being interacting with it could tell whether or not they were responding to another human being or to a computer. It turned out they couldn't. That was more than half a century ago, before the phrase Artificial Intelligence came into common use. We now live in a world where we are told we can soon expect to have or already do have machines that can not only emulate or be a human mind but can provide convincing companionship, medical and domestic care, and even physical love. We are even being warned that AI can take over, become our masters, make us irrelevant and obsolete.

Human beings are hard-wired (note the machine metaphor) from infancy to recognize and distinguish human faces. Other mammals are similarly constructed. Elephants can distinguish, recognize, and remember smells better than any canine. It's an essential ability for any species to be able to tell friend from foe, family from stranger. We humans have that ability to such a degree that we tend to see faces (and other figures) in clouds, in the surface of the moon, in folds of clothing, even in a grilled cheese sandwich or a strangely-shaped vegetable. In other words, we seem to be made in such a way as to see random shapes as human forms even when our reason tells us otherwise.

The same proclivity to anthropomorphize seems to take over when we talk to a machine programmed to talk back to us. The software engineers who create these programs use anthropomorphic language to describe what the machine does, freely admitting such words are only metaphorical. But despite their warning, being a literal-minded species, we assume the machines are doing what they seem to be doing: thinking, remembering, and most important, experiencing what we humans experience as consciousness and even personhood.

We got ourselves into this scary situation by not recognizing that what it really means to be us, indeed to be any kind of living being, consists in not just what we recognize as thought and consciousness but in the sum total of our entire selves from toenails to scalp. We assume, as we have for the entire history of Western civilization, that we are non-material minds residing in material forms we call bodies. Even those of use who do not share the religious belief that the physical world is inferior to a higher, non-material one still assume our bodies are just a kind of support system for our real selves, which may on that account survive our physical death. We even like to believe that thought and what we call emotion are two different things, one being of a higher, more properly human order, the other being a vestigial part of our older, pre-human brains.

We don't, or prefer not to, recognize that we think with our whole body, guts and all, not just with our gray matter. And by "think," I mean something broader than rational thought or even emotional experience. It's not despite but because we experience discomfort and release in the forms of hunger and satiation, pain and relief, itching and scratching, defecation and urination, as well as intense pleasurable experiences like orgasm, that we can also experience moments we associate with non-physical pleasures like music and logical thought, dark nights of the soul and supernatural bliss, the recovery of an elusive memory, phobia, joy... and everything else we experience during our waking and even non-waking hours.

We have been taught, even if not directly, that our "higher" cerebral forms of experience are different from those we have when we are tasting our favorite ice cream or venting an overfull bladder. For whatever the reasons, we tend to be uncomfortable with the way we are actually made and reject or at least demean the "lower," strictly "physical" part of our selves. Of course, cultures vary a great deal in how schizoid they are on this account. And animals in their wisdom don't pretend their feelings of pain and of love are located in different faculties from where their consciousness resides. They simply get on with living, and do so without the self-reflective and largely useless navel-gazing in which we engage.

It's difficult even to write about this subject because the vocabulary we have for it prejudices the way we see it. That's true for anything we can discuss or think about, but in most other matters we usually have some recognition of that limitation. When it comes to Artificial Intelligence, we are so in awe, so full of fear and expectation and, most importantly, so predisposed to anthropomorphize everything the cyberneticians report to us on the subject, that we lose perspective.

But we do the same thing when presented with new information about other sciences. Think of the misconceptions we keep making about genetics despite the cautions we hear from those in the field against oversimplifying and jumping to conclusions. Do you know anybody, no matter how low or high their position in society or educational credentials, who does not still use the scientifically discredited and dangerous term "race" in both private and public discourse? Why? It's because we have no alternative term for the concept. And why is that? Because the notion is a fixed idea that seems obvious and inescapable because we absorb it with our mother's milk, though other cultures find it bizarre, having no equivalent to our own "one drop" rule.

For a machine to be conscious or to "experience" anything as a living being, it must have the complete form of that being, its flesh and bones and organs, not just enough electronic hardware and software to give an illusion of personhood to a gullible species like ourselves. We think, have consciousness, with our hands and feet, guts and muscles, just as trees "think" with their roots and leaves. To be sure, there is a way to make a machine identical to all the functions we have as human beings. The polite term for it is sexual intercourse.

If there is a stumbling block to our understanding this, it may be the extent to which we don't appreciate a couple of facts of life. The first is the extent to which everything we think we "know" is actually a creation of our brain, a modeling that goes on 24-7 and without which we know and experience nothing. The second is that this constant creation or modeling going on not just in us but in every living thing must be as much a force of the universe as is gravity or any other so-called physical law.

We do not experience anything directly except our own experience. To the extent Descartes's "I think, therefore I am" has any validity, that's it: I experience myself thinking, therefore the self doing the experiencing must exist. Again, this is true for any living creature. The corollary, free will, has validity for the same reason: I experience choice, therefore I must have such an ability. Again, this is no different from what a poodle or a dragonfly assumes.

But that experience or apprehension is a creation of our own brains out of the mix of sense data we are constantly bombarded with both from outside and inside our skin, not a direct apprehension of anything at all. Living, being alive, is a giant act of faith in the reality of this massive creative process which all of life is constantly engaged in just to keep itself nourished, never mind figure out the value of pi or paint the Milk Maid. Light and soundwaves have to be transformed into electrical currents, out of which our gray matter makes a reality, from a pretty face to a tasty bit of plankton. We live by faith in this model we call reality, literally and cognitively live by and in it. But it is a creation, not an objectivity.

The idea we live in a kind of virtual reality can be depressing, even claustrophobic. Are we trapped inside an artificial intelligence of our own making that is no more in touch with the real world than is the cold metal and silicon of our laptops? But such a realization is only depressing if we insist on our being Godlike beings who can apprehend things as they truly are. If we accept that everything-as-it-truly-is does not exist, that there is no alternative to the creation we ourselves make, which the cosmos itself makes through us via a force as natural and inevitable as gravity, it doesn't seem so constricting or, to be blunt, humiliating. Quite the opposite. If what I am doing writing these words is the cosmos creating itself, that's a job description I can live with.

A young John Stuart Mill fell into a deep depression when he realized the beautiful sunset he was observing was the result of particulate matter in the atmosphere. He got over that depression and learned to enjoy life again. We can, too. If there's no "objective" truth, i.e., no reality we can apprehend as-it-is, stuck as we are in an ocean of artificiality, we can either accept the wonderful fantasy world we create or whine about not being able to do the impossible. But that's not a state of mind you'll find any sparrow succumbing to on a sunny day, or any happy child, either. If we must live by faith, with only brief, metaphorical insights into a scheme of things we can not and never need experience first-hand, we should learn to enjoy it and not worry about machines becoming the masters we ourselves were never meant to be.