The world was startled in the past few months by the abilities now achieved by artificial intelligence. Chatbot programs can write competent sonnets on command and whole essays on almost any conceivable topic. This is an advance over the wonder expressed previously over computer supremacy at mere chess. Immediately, the public asks when these robots will qualify as humans dictating affairs, which in turn, raises the profound question of what in fact constitutes humanity or is essential to it. What must we know and protect?
Command of language has long been cited as a distinctive capacity of human beings, but those who are intent on narrowing the distance between man and lesser animals, let alone machines, have long made a game of demonstrating that donkeys can count, and monkeys can master sign language. Whales are intelligent; dogs have emotions. It’s true that the human mastery of symbolic forms is distinctive in this universe, and we are adept at communication, but it is noted, animals, too, communicate. So maybe, as with animals, it is just a matter of degree between people and educated machines, with no qualitative differentiation. When the AI robots equal us in intelligence or language, will they then be “human”?
A Brown University professor of computer science, Michael Littman, was on the affirmative side in an interview in AI magazine: “After all, humans are just a particular kind of machine.” The rise of AI brings to the fore the fundamental matter of the definition of the human, and the nature of the qualitative divide between man and machine.
A recent essay by John O. McGinnis settled on the possession of morality, the inescapable sense of right and wrong, as the essential and unique human attribute. Maybe dogs feel shame and elephants are loyal, but that is hardly a code of ethics, let alone an injunction such as “do as you would be done by” or Kant’s moral imperative: “Act only on that maxim that you can consistently will to be a universal law.” Can robots be trained to make moral distinctions? No doubt they can and will be, after ingesting tens of thousands of examples of right and wrong. But even with this, won’t they still lack sentiment? Will we be able to appeal for mercy to a robot judge in a courtroom? Will our judge feel compassion? Can a data-based government offer clemency? Can a robot exercise judgment as opposed to calculation?
These are disturbing, even frightening, matters because we are all aware of a scientific mindset that thinks superficially only of mechanical advances not of tenderness and sympathy in the human realm. Will the vibrations (or call it communication) of a human mother singing a lullaby to her baby be superseded by a robot?
Thou and I
The brilliant logician Alan Turing, beginning in the 1950s, famously proposed that we would know when robots were human-like, that is, can think and be intelligent when we can converse with them naturally to the point that we will be unaware we are in fact conversing with a machine. Long ago he confidently foresaw huge advances in machine learning, and with powerful reasoning he directly refuted skeptics. His emphasis, however, was on conversation, informal language, or even professional discussion, but not, it must be noted, on what the historian and social philosopher Eugen Rosenstock-Huessy (1888–1973) called speech.
Rosenstock-Huessy could hardly have imagined the technological ingenuity that has brought us to this moment of confusion. He commented simply on what constitutes the essentially human, on what is at the roots of human society, although that was not his central theme. Conversation, chatter, talk, language, communication, and philosophical and technical discussion are all at a lower level than speech because the talk is not binding, and such spoken words, mere words, do not demand a response or a commitment. “I only begin to speak,” Rosenstock-Huessy said, “when I submit myself to my own words.”
Speech between humans requires, first of all, personal names, the distinctive attribute we are given at birth that identifies us singularly throughout our entire life. Names also extract us from the so-called natural world. Personhood is conferred by naming and then by direct address: Hello, infant Mary or infant John; welcome to the world of people––not just of material things, not just of all living creatures, but to the company of humanity specifically.
Rosenstock-Huessy pointed to the well-known fact that compared to the development at birth of other mammals, humans come into the world totally helpless, about nine months prematurely by comparison. Those first nine months out of the physical womb he called the socializing womb, the womb of speech, when we are constantly addressed by name and called into being, as one might say. It is the constant addressing of Mary or John that confers upon them an awareness of self––I am Mary, I am John––that is, the transition from being addressed as “you” or “thou” to the awakening of Mary’s and John’s responsive “I.” We are commanded in infancy and childhood not only to “eat,” to “say thank you,” to “go to sleep,” but also to “come forth,” to “be.” The circumstances of our early childhood give us, first, a past and present, that is, a surrounding bed of nurturing, and then an awareness of another dimension of time, the future, the image of our individual named selves becoming (or, more elaborately, becoming something in the world, an expectation).
Morality may be essential to humanity, as McGinnis observed, but it is grounded in a sense of self, and that selfhood was brought into being initially by direct address: you Mary, you John, which is then followed by their “I shall.” Everyone is familiar with Martin Buber’s famous book I and Thou. Rosenstock-Huessy, who was a lifelong friend of Buber’s and himself a premier dialogical thinker, criticized the Buber formulation on the grounds that empirically “thou” in life comes before “I”; that is, the second-person in grammar, defined as the imperative––[You:] Go; [You:] Listen––comes first in human development and human affairs, followed by the first-person “I” response. I and Thou should be more accurately formulated Thou and I. This process of call and response continues throughout life.
Machines may have clocks, atomic clocks, but can they anxiously or hopefully anticipate or fear the future? Will they be able to reflect upon their pasts, with both joyful and regretful memories? Can they know they will die and “live their lives,” so to speak, with that acute awareness? Time is of the essence. For Rosenstock-Huessy we may occupy space, but we live in the midst of time. For humans, time is not the physicists’ fourth dimension of space; it has its own far more vital three dimensions of past, present, and future. “Man is peculiarly a temporal being,” Rosenstock-Huessy said, “ever but an exile and a pilgrim in the world of space.”
Because of the relentless undifferentiated flow of the physicists’ time, we humans lock in moments to humanize that stream. We celebrate birthdays and anniversaries, we establish national holidays and family holidays, and we have calendars that keep track of where we are in the year two-thousand-and-twenty-three. A couple of millennia is obviously just a blip in geological let alone astronomical time, but this is our common (formerly Christian) era, distinguished from meaningless tick-tock, tick-tock. Machines can dazzlingly count and calculate, but will they be prone to celebrate meaningfully, to commemorate? “Men conquered time when they began to speak,” Rosenstock-Huessy reminded us.