Black Robe and the Rise of the Machine

In the 1991 film Black Robe1, a Jesuit priest sits on a log in the forest, writing in a book. It’s Quebec in the 17th century2, and he’s travelling with a group of Huron aboriginal people and French explorers.

A Huron man asks the priest what he is doing.

“Making words,” says the priest.

“Words? You not speak?”

“I will show you something,” the priest says. “Tell me something I do not know.”

The Huron thinks, then tells the priest that in the previous winter, his wife’s mother died in the snow.

The priest writes the words in his book and then walks over to another French man in the company. He asks him to read the words aloud.

Words of a trickster

When the priest’s companion reads the words aloud, the Huron comprehends. They are jolted. How could this stranger utter the words he had just said to the priest on the other side of the campsite? There was no written language among the First Nations people. The Hurons understood the power of the written word for the first time.

When the Hurons are by themselves again, one mutters in his own language (via subtitles):

“He is a demon!”

They see the priest as a trickster and are deeply suspicious (and well, they should be, considering what happens in the movie).

Things don’t work out well for the native people in this film (nor for the Jesuit/church people, for that matter). But their exposure to written language and European civilization in this scene could have changed their world.

The introduction of the written word may be one of the early “technological advances” in the world, like the steam engine, human flight or the Internet.

Is Artificial Intelligence the new ‘demon’?

Today, we’re living through an explosion of Artificial Intelligence advances. ChatGPT allows you to have intelligent and wide-ranging conversations with a computer. It can help you compose emails, essays and make business decisions. Doctors and lawyers use AI, and practitioners in technical fields. We’re just scratching the surface.

It’s scary, disturbing! It seems to be moving too fast! What does it mean for the future … for employment, society, morality, politics, and art — all of which AI has touched. Where does it end? Is the genie out of the bottle? What will happen to humanity?

I feel a little like the Huron man in Black Robe. Is AI the new demon?

No one knows where we’re headed with artificial intelligence. We have a sense of what it promises, along with a host of real or exaggerated fears that could lead to our doom.

“Surprisingly strong and getting stronger …”

In a 2017 documentary, AlphaGo3 Korean Go champion Lee Sedol speaks confidently at a press conference. It’s the eve of his first game facing the new “artificial intelligence” machine developed by London-based Google DeepMind to play the game of Go. The developers had named the AI AlphaGo.

Go is an abstract board game for two players, much like chess. Players use black and white stones and place them on vacant intersections of a 19×19 square grid. As the players play the stones, the goal is to gain the largest territory on the board by surrounding an opponent or region. The player with the most territory at the end of the game wins.

Go was invented in China more than 2,500 years ago and is believed to be the oldest board game continuously played today.

It’s an incredibly complex game! One Go expert claimed there were more board configurations than atoms in the universe. I’ve never played Go, but I have watched a few games. Frankly, it’s beyond my comprehension! But you don’t need to understand the game to appreciate the movie, which focuses on the development of Go-playing AI.

“I have heard that Google DeepMind’s AI is surprisingly strong and getting stronger,” Lee Sedol says at the press conference. “But I am confident I can win at least this time,” he said.

Playing for all of humanity!

“I am not just playing for myself and my country, but for all of humanity. I am going to do my best to protect human intelligence.”

Lee Sedol has the face of a young boy but speaks with confidence and poise. He’s a likeable fellow, and his faint smile underlines his assurance that he will succeed. He predicts he’ll be the victor in all five scheduled matches.

But we in the movie audience suspect things might not be easy for Lee Sedol. This is another story of human vs. machine.

A shocking move

The competition goes badly for Lee Sedol, who is playing with the black stones. He loses four of the five games. In the second match, the computer made a play that left everyone in the audience breathless — Go experts, members of the general public and AI developers themselves. It was at a critical point of the game, move 37, to be exact. Most thought the play was a mistake.

Lee was taking a break when the move was made. On his return, he sat down and beheld the board. He winced! You can see dark clouds gathering around him. He fidgets; his brow is tense. He hesitates, makes several false starts, then places his countering move. But the course is set, and he can’t recover the advantage he thought he had.

Meanwhile, the commentators and AI developers had come around to the move and saw its genius. One said, “Not a single human player would have chosen move 37!”

Lee is apologetic to the audience afterwards but gracious in defeat. He congratulates the DeepMind team that developed the AI and vows to play better next time. He’s shaken, near tears, it seems.

A turning point

Watching the film, it felt as if the match was a turning point in the development of artificial intelligence machines: a turning point for the AI researchers, but also, I feel, a turning point for humanity.

I can’t say if AlphaGo is true artificial intelligence. We’d have to let the experts answer that. But the achievement of developing a program to play the most complex of human games and beat one of the most advanced players is staggering. The genie is out of the bottle.

Lee won a single game in the five-game contest. It was a narrow victory and a relief to the Go world that he won at all.

In 2019, three years after his loss to AlphaGo, Lee Sedol retired from professional play.

“With the debut of AI in Go games,” Lee said, “I realize that I’m not at the top, even if I become number one through frantic efforts. Even if I become the number one, there is an entity that cannot be defeated.”

“An entity that cannot be defeated”

Subsequent versions of AlphaGo have gotten much better. Developers produced a new iteration, in 2017 that contains no “human datasets.” Once the program (dubbed AlphaGo Zero) learned the rules of Go, it was on its own, teaching itself to play and learning from every game it played.

Finally, the AI developers tested AlphaGo Zero in a tournament against the earlier version of AlphaGo, the computer that defeated Lee Sedol. Two computers battling for Go supremacy! Of the 100 games played, AlphaGo Zero won them all. 100 to 0! Now that’s a shutout, a bittersweet result to most.

The machines may finally be gaining a foothold in the order of things, but something is appealing for optimism in this story!

What I’d like to believe is that it wasn’t necessarily AlphaGo’s intelligence that allowed it to succeed. It was the fact that it learned the game from scratch — a tabula rasa, or clean slate. Developers provided no human input, lessons or coaching, or reliance on expert opinion or experience. AlphaGo followed the game’s rules without the accepted cultural norms and narratives gathered over millennia. In other words, it wasn’t held back by limiting beliefs.4.

The full spectrum of possibilities

The AlphaGo vs. Lee Sedol tournament wasn’t just a turning point in AI development. It was the first time Go was played with the full spectrum of possibilities. With fresh eyes, so to speak, the AI was able to innovate, devise something completely new, and transform the game forever. If it had been taught to play by humans, it might not have won the match.

By teaching the AI to learn the game independently, programmers may have accidentally given the program the advantage that led to move 37 of the second game. Experts and observers initially called the move a mistake! And gradually, they recognized a computer-driven performance that none had expected or seen before.

“I thought AlphaGo was based on probability calculations and merely a machine,” said one observer of the historic game. “But when I saw this move, I changed my mind. Surely AlphaGo is creative. The move was really creative and beautiful.”

A Chinese analyst said: “This move is very special because, with this move, all the stones played before now work together. They are all connected. It looked like a network; they linked everywhere. It’s very special.”

Is there a “move 37” waiting to be played elsewhere?

If AlphaGo could produce such a radical move, other AI machines could be used to solve different problems. Climate change, global conflict, poverty, famine, etc. This story suggests that human solutions to problems, while substantial and helpful in many ways, may also be limiting and destructive in other ways.

Maybe an AI “move 37” is waiting to be discovered in any number of problems facing the planet. Solutions that may at first look like mistakes but, in fact, are creative, beautiful, powerful and effective.

Maybe we humans need to get out of the way and let the machines rise. Teach them a few rules, then let them change the world.

Does this sound too good to be true? Maybe, but what have we got to lose?

---

---