Forbidden Fruit

a.i. bites off more than we can chew

In today’s computer age, there is a theory among the chess community that the game is dying: The current world champion doesn’t stand a chance against a chess engine installed on a smartphone, and everyday computers map out corners of the game’s theory that no grandmaster a century ago could comprehend. The most powerful supercomputers have already solved every possible game of chess with any combination of seven or fewer pieces on the board, and they are currently approaching eight. There’s a saying in chess that there are more possible moves in those 64 squares than atoms in the universe, but it seems now that, one atom at a time, that universe is being drained of its mystery.

I have always been addicted to chess’s artistic side, the part of the game that goes beyond cold calculation and melds with creativity and beauty. While it is true that computers have mapped out huge portions of chess, I’ve always resisted seeing computers as the death of the game, because they have also produced some of the most inexplicably breathtaking moves I have ever seen.

I remember spending one morning in 2014 studying the Albin Counter-Gambit, a chess opening setup regarded for over a century as a third-tier amatuer choice. I was running Stockfish, the strongest chess computer program in the world at the time, as my “assistant” on my laptop. I forgot to turn Stockfish off when I was done, and it turned out to be one of the best mistakes I have ever made. When I returned a few hours later, the monitor displayed something mystifying: In a chess position played over in hundreds of games and analyzed in articles around the world, Stockfish was suggesting a shocking move never seen before. It was a bizarre idea, seemingly suicidal; and yet when I delved into Stockfish’s thoughts, I realized that I was staring at a revolutionary find, one that could breathe new life into this opening. The lines of moves that Stockfish suggested were outlandish and otherworldly, utterly impressionistic and beautiful on the outside, but devastatingly powerful due to its brutal calculating vision. Stockfish suggested a loss that was not only considered fatal in chess, but would also put the opposing player’s position in complete paralysis, a concept never seen before in any books or archives. In that moment, I saw the magic of Stockfish, the lone creator and destroyer of chess’ future.

A few weeks ago, a piece of news appeared that shook the whole world: AlphaZero, the machine learning entity created by DeepMind, had obliterated Stockfish in every game out of the 28 it played. AlphaZero had taken just four hours to master the game from scratch. This was incomprehensible; Stockfish was supposed to be supreme, unbeatable—as close to chess perfection as we were ever going to manage. What was AlphaZero seeing that Stockfish, a program that can calculate millions of moves every second, could not? Yet, this was not even where the true shock factor lay. DeepMind also released 10 games from the match. Looking at them for the first time, I have to admit that I did not know what to think or say. They were utterly alien. It was as if AlphaZero were on the other side of some all-knowing veil, standing on some new plane of chess understanding. I felt like I was staring at the work of a god. There was one final nail in the coffin—in all 10 games, Stockfish did not stand a chance. Chess computers are generally boring in their style and stingy with their pieces, preferring to suffocate their opponents to death with their “perfect” play. AlphaZero’s style, on the other hand, was sleek, turbulent, almost human. In some games, it had a bloodlust and thirst for direct attacks more severe than any human could muster, sacrificing pieces left and right with cold rage and controlled recklessness. In other games, it played like a python, strangling the life out of Stockfish with a precision that was agonizing to watch.

One might think that I would be awash in appreciation for AlphaZero’s ingenuity, having seen a new level of creativity in chess, and to an extent this was true. My jaw dropped with each game I went over, and the warmth and excitement of falling in love with a game of infinite mystery reignited in my senses. Yet, deep in my chest, I felt something was off. I felt fear.

It took four hours for AlphaZero to surpass over a millennia of chess knowledge. There was something so impressive, yet so hollowing in this fact. All the chess greats, all the “masterpieces” I looked up to with all my heart growing up, seemed now like flies circling a blinding light. For an instant, the chessboard was no longer a place of mastery or a blank canvas to me—it was a void, and I had just met the entity that knew everything about it. That feeling, of staring straight into a dark expanse filled with so much wonder and unease—I felt like a child again, outclassed, outmatched, and utterly helpless.

As students here at Brown, we’re growing up in a world that is changing faster than ever before. The question ringing in people’s minds around the world is no doubt how to face it. Yet, in the midst of smog-breathing cities and an economy teetering on excessive spending, we continue to embrace the same selfish thought as every generation that came before us: No matter what evils or advances are embraced in the future, we will likely not see it—how fast can the world change in 80 years anyway? The answer to that question now cuts me to the core some days. We will be the first generation to create some of the most daring and groundbreaking advancements in technology, and we will also be the generation here to see their consequences.

I’ve always been a sentimental person with what I’d like to think of as a hopeful, even romantic heart. I wish I could look at this in an optimistic light and say that the future is brighter than ever before; I truly hope that technology like AlphaZero will be used to create the transcendent ideas I’ve always dreamed of in chess, and that whatever AI algorithm DeepMind invents will be used to provide food, energy, and livelihood with greater efficiency than ever before. But the truth of the matter is that my heart is tainted with fear. The world is filled with selfish incentives and unchecked recklessness. The race now between the superpowers of the world to invest in AI reminds me of the time we discovered the unholy energy generated by splitting a uranium-235 atom. The children of those scientists discovered even more cataclysmic ways to kill, and now every day we hover our fingers over launch buttons, fearful that our time of reckoning has finally come. Perhaps the scariest thing of all is that we can’t imagine what weapon or fallout is possible with AI—only they can. If a current prototype could master one of the most complex games on earth in four hours, future intelligence will have a relentlessness that we are not ready for. I can’t help but wonder if some things are not meant to be discovered; we push and keep pushing every day, forgetting that we may be eating forbidden fruit.

I’m becoming scared of the things I don’t know in this world. Pride has given way to an almost crippling open-mindedness, an alarming inability to choose between what I think is right or wrong, in a world where that choice is more clouded and needed than ever. I’m at the age now to decide what to believe in, what values to cling to when the future becomes critical, and I’m bombarded with how I should think every day. My own compass is now pointed in jagged directions.

Today the world is developing almost like a computer would, choosing efficiency and speed over livelihood, and obsessing over large power plays while forgetting the millions of individuals that are the pawns of each drastic decision. In an age of globalization and stunning technological advances, we are at our most empty in our humanity. We are trained in cold cost-benefit analysis, learning how to extract every drop out of this world where the winners and heartless take all, a world filled with soft things and easy pleasures. As the future and all its terrifying possibilities collide, we must remember not to push the boundaries of comfort and power too far, to not try to escape the ailments of our flesh and blood—the most poignant reminder of what makes us human.