The limits of artificial intelligence

We should bank on human intelligence, and invest in types of education that will develop the human capacity to understand

This is speechmaking season. I have found myself in front of audiences for graduations, retirements and award celebrations. All this speechifying has taught me, forcefully, the limits of machine learning and artificial intelligence. I have been giving what used to be called “occasional speeches”. This does not mean, as some might hope, that they happen only occasionally, but rather that their purpose is to respond to an occasion or event, to meet and match a moment.

What kind of moments have I encountered? These bright-eyed, or maybe bleary-eyed, kids are about to leave a sheltered environment and enter the world, now at last pursuing happiness fully guided by their own lights. What kind of world will they find? What tests and trials will they discover?

Or: This man is retiring. On the basis of what kind of performance? What life has he crafted? What rewards have been earned by the worry lines deeply etched in his brow, and by the laugh crinkles beside his eyes?

To meet and match these moments, a speaker has to register the wants and needs, interests and motivations, hopes and fears that have brought everyone into the room, or onto that open campus greenway. This family member has never been to a college graduation before; that one has generations of graduation pictures on the walls of the study at home. This one is struggling with anxiety and barely even made it to the graduation ceremony with clothes in order and tickets in the pocket or purse. That one has been working extra shifts steadily to achieve an oasis of security. All of them wonder what to make of the turbulence of the world. All of them want to find a way to do right for their loved ones — to find equilibrium — despite the gyrations and confabulations of famous and powerful people.

But how to name the moment and point to a worthy path? This is the challenge.

To master it, a speaker must understand and answer the question presented by a roomful, or a field full, of upturned faces. The relevant question arises from the synthesised aggregate of the life experiences, sources of pride, hopes and fears, questions and concerns brought to the fore by the occasion. What human question is on the table at this time in this place? On this occasion?

What we often call the vibe in the room is that human question, pulsing.

In a few quick minutes, in registering that vibe, human beings process an infinitude of data — emotional, social, experiential, political, broadly contextual. We read the room, as they say, like a book, something we learn how to do well at least in part by reading books. Literature, history, philosophy.

Even our fastest machine processors, fed by our most able data aggregators and structures, taught by our best algorithm writers and then by themselves, cannot process this much data of this kind this fast. Our brains own this territory.

What is at stake in this synthetic function of human cognition and understanding? I’ll put it out there, grandiose though it may sound. The shape of our world is at stake. Not in this or that commencement speech, but in all occasions taken together when human beings speak to one another with the goal of naming where we find ourselves and where we should be going.

A good speaker grasps the question on the table for her audience and offers the best answer she can. If she is so lucky as to succeed, she sends back into the world a group of human beings whose own mental furniture has been slightly, even if only slightly, rearranged.

If and when, through the contributions of many thinkers and talkers, these rearrangements add up to a full shift of human perspective, human culture begins to shift gently. Sometimes, if rarely, one speech, on the right occasion, can quickly and radically establish a new direction that others follow up with long labourious work. (My speeches don’t achieve this. But we can all think of our favourite examples; the Rev. Martin Luther King Jr. always comes to mind.)

Artificial intelligence is transforming our economy, but is light years from being able to guide our culture. It can navigate our cars, but not us. And, of course, only we can choose the destination. For both of these things, I’m still placing my biggest bets on human intelligence.

I’m therefore investing in forms of education that develop our distinctively human capacity to understand and meet our moment and to do so with honesty and integrity. Perhaps I’m in the minority, but I’m still betting on teaching and learning that fully incorporates the humanities.