EDITOR’S NOTE | By Bret Bradigan
A.I. in Ojai?
It’s a busy weekend in Ojai — besides hosting America’s most wonderfully hokey parade, we’ve got art openings and live music and retreats and full hotels. Our streets brim with life in all its technicolor glory. The pageantry of humanity is on full display in our beloved “Little Orange” this week.
There is something poignant and sentimental about humanity — our fragile mortality, our outsized dreams and ambitions, inescapable weaknesses and surprising resilience and the power of our love to bend the arc of history toward justice. It seems absurd that we could ever be replaced by machines.
As a writer, I have a vested interest in the peculiarities of human intelligence, our ability to make connections between disparate elements, thereby bringing a deeper level of meaning to the human condition.
It’s hard for me to conceive that artificial intelligence could one day replace writers in their very human struggle to wrest meaning from the world’s randomness. I was wrong, and that “one day” may already be here.
Back in the 1990s, an early machine learning program could take a scoresheet from a baseball game and within seconds, create a persuasive account of the game – mix in some concrete nouns here and some action verbs there, toss in an off-the-wall metaphor every now and again, and it would indistinguishable from the work of your typical sports reporter.
As an occasional sportswriter myself, I could see the writing on the wall, so to speak. Now, that program never did live up to its promise to put sportswriters out of business and save publishers dozens of dollars (sportswriters being notoriously underpaid,) but machine learning has entered a whole new ballgame.
Last year an A.I. program called AlphaZero took about two hours to beat the world’s best chess players, both humans and computers. By comparison, it took years for IBM’s Deep Blue to attain a mastery sufficient to beat Gary Kasparov, the world’s best in 1997, and only with massive amounts of data about every single game of chess for which moves were remembered. AlphaZero, on the other hand, was programmed only with the basic rules of the game. Its secret sauce is its ability to make unexpected connections, learning as it goes at lightning speed.
In 2016, AlphaZero’s earlier incarnation, AlphaGo, defeated Lee Sedol, the top player of Go, considered an infinitely complex game that could only be mastered by humans and our supple, creative thinking. The students of the game were wrong then. And have only gotten more wrong since then.
By 2025, A.I. is expected to replace 16 percent of jobs in the United States; disrupting our transportation and manufacturing industries beyond recognition, leaving massive disruption in its wake. That’s just the first wave.
A lot of people feel that writers are immune to this wave of disruption, because, according to the popular blog, “Pro Writing Aid,” “What artificial intelligence and machine learning are missing is the art of narrative or storytelling. We humans are emotional creatures. Writers appeal to fear, joy, love, persuasion, anger, and a host of other human emotions when creating content. And writers of all forms understand that storytelling is important because it helps other remember what we tell them.”
But that’s assuming the “art of narrative” is something that can’t be learned, that it is a uniquely human form of expression. Those are basically the same arguments made by Go and chess players. How could a machine learn the creative spark of connection that identifies us as human?
Maybe machines can’t. But then again, maybe it doesn’t matter. Because the disruptions would be so wide and so deep, it may be impossible for humanity to keep pace. That’s why I’m gradually coming around to the notion, promoted by Andrew Yang’s presidential bid, that universal basic income is a good idea. It may be necessary to keep civilization’s wheels from coming off. What are we going to do with all these surplus workers?
Perhaps the subtleties of the human experience are far too nuanced, too shaded with uncertainty, to ever be duplicated, but maybe they aren’t. Maybe even if it’s mimicry, it would be so advanced that no one would ever be able to tell the difference. At that point, we are inside the Matrix, everything such a convincing simulation that the lines between real and artificial aren’t blurred, but obliterated.
As Yuval Noah Hariri points out in “Homo Deus,” we should be kind to all living creatures as well as to our machines, because a big part of machine learning is copying our behaviors. If they see us as cruel – as we demonstrably are when it comes to our treatment of livestock and fowl — that’s likely to be their default setting. And if we are kind and tolerant, if we demonstrate the better angels of our nature, so Hariri’s thinking goes, then so will they.
Compassion and kindness would be a fitting memorial to the human experiment, to our ascent from the savanna of the Great Rift Valley to the stars. If we can leave that as our legacy, then we have triumphed.
In the meantime, I’ve got to go. There’s a fireworks celebration to attend with several thousand of my closest friends.