Deep Blue, AI, trouble ahead

Kasparov's loss to Deep Blue in 1996 set the stage

World Chess Champion Grandmaster Garry Kasparov lost to computer Deep Blue in 1997

Reading my friend Ronald Gruner’s excellent piece on AI (recommended!) reminded me of the great Garry Kasparov vs. Deep Blue computer chess rematch showdown in 1997 and how it truly foreshadowed what we’re seeing now with AI everywhere you look. If you haven’t noticed, AI is just out of control; there’s a long, scary story in the New Yorker about how it is impacting college professors and cheating students. Did you write this or did AI? (FYI: I writed this. And I always will.)

In 1996, Kasparov, a brilliant chess champion who was then thought of as the greatest player in history (Magnus Carlson probably holds that title now) had played chess computers before with great success. While a chess-playing computer can calculate millions of moves which a human brain cannot, a chess computer does not sequence its moves — in other words, it doesn’t do Move C because Move B and Move A set things up. The easiest way to explain it is, every time a chess computer looks at a position on the board, it’s as if it’s seeing it for the first time. “Oh, this is the best move” regardless of what move came before. Does that make sense?

Kasparov just attacked wildly and the computer was overwhelmed and easily lost. In the 1997 rematch, however, Deep Blue’s computer firepower was greatly enhanced, there was even more publicity and it seemed Kasparov’s brain was standing up for all mankind.

In the rematch, Deep Blue’s on-the-board play was considerably improved. Kasparov noticed it right away. Since the IBM corporation was shamed by the first loss, Kasparov suspected they had hired a team of Russian grandmasters to advise Deep Blue on every move in the rematch - IN EACH GAME.

Conspiracy goes hand in hand with Russian chess. The great American champion Bobby Fischer, for example, charged that in world tournaments on his way to challenging for the world title which he finally won in 1972 from Boris Spassky (the title was always owned by Russians) ,other Russians would simply play for draws against one another to make it that much more challenging for a non-Russian player (i.e. Fischer) to advance in the tournament. And he was probably right.

In playing against computers, one tell-tale sign is, generally speaking, a computer will always take material (i.e.: pawns, rooks) over position. There are times in a chess game where you can sacrifice a piece to gain a strategic positional advantage. A computer would not do that. Ever.

So, in the rematch with Deep Blue this time, Kasparov got in a position where he offered a piece to Deep Blue and IT DIDN’T TAKE IT. That set him off. Then, a bit later, the computer made a move that just BLEW HIS MIND. A sacrifice of sorts, a strategic move the likes of which he had never seen a computer make. And it shook him so, the conspiratorial air all around the match just intensified and he lost. And so did mankind.

In checking the computer program after the match and studying the process for that particular out-of-character move, in ironic fashion, IT WAS A GAFFE! The computer made a very human-like mistake and made a strange, random move. Which convinced Kasparov that something fishy was up and he fell apart.

So a man lost to a computer because the computer acted human — and made a mistake. Amazing.

Some 28 years later, computer technology is doing everything you can think of and some you can’t. In colleges, professors are stumped as to how to deal with AI, ChatGPT doing the writing for their students. In a recent New Yorker article, a survey of college and university professors reported a 59% increase in cheating, “a figure that is conversative,” the magazine says, “when you talk to students. AI has returned us to the question of what that point of higher education is.”

“One study, published last year,” the article continued “found that 58% of students at two Midwestern universities had so much trouble interpreting the opening paragraphs of “Bleak House” by Charles Dickens, that ‘they would not be able to read the novel on their own.’ And they were English majors.”

Dickens’ opening to “Bleak House” was revolutionary. I wrote about it a while back. But English majors shouldn’t have been thrown by it, should they?

When Garry Kasparov lost that chess match to Deep Blue in 1997, some people speculated the human brain would someday be obsolete. I’ve read that AI is writing novels, now, doing all sorts of things previously only done by humans.

Fortunately for me, I got out of the teaching business just as this AI stuff was arriving. I had enough trouble back then with students plagiarizing stuff off the Internet —ME: “What does dystopian mean?” STUDENT: “Uh” — I can’t imagine having to wade through paper after paper, wondering who or what actually wrote it.

That’s progress, I guess. Or is it?

HERE’S WHAT I WROTE ABOUT “BLEAK HOUSE”

https://johnnogowski.substack.com/p/did-dickens-recognize-his-breakthrough

HERE’S A TIME MAGAZINE STORY ABOUT THE MATCH

https://time.com/3705316/deep-blue-kasparov/

HERE’S A FABULOUS DOC ON THE KASPAROV-DEEP BLUE MATCH