Okay, I tacked that last bit on at the last minute but it kinda fits. This post was spurred by a video by
Levy over at GothamChess (on YouTube) – Chess is Dead. I’ll spare you
detailing his rant. Entertaining as it was, his premise was based on a variation of a chess opening played by top GM Magnus Carlsen. The variation and the derivatives* of it have spread over the the internet like wildfire.
* FINALLY! I think I’ve stumbled upon a somewhat clear depiction of derivatives. Maybe I’ll post on that in the near future.
The bottom line is Levy’s contention that the sphere of chess knowledge is becoming finite and thus relegated to rote memory. HUH!?
Excuse me while I yawn. Okay, I’m back. Sorry about that. Am I bored? Not really, but the fact of the matter is, I actually anticipated all this decades ago, though I didn’t realize the implications of those thoughts. Let’s back up here while I explain and hopefully clarify all of it.
Years ago I had an idea for a story. It was just an idea and I never developed it. This was back when I worked for a living and wrote now and then. These days I only work when I must and I write now and then. Anyway, the story…
It opened with a chess move. The verbal response was not a move but a statement:
“You’ve lost, Ben.”
“What?” our hero Ben replies, “I only made one move, how can I lose already?”
“I’ve calculated the result. There is now way you can win in this position”
Now you tell me – isn’t this exactly what Levy declares we on the brink of today? No matter what anyone does, the results have already been calculated and therefore there’s no point in playing chess anymore. So, a couple of thoughts here.
First up, this may be well and good for all those top players out there, but for me? No. I’m not about to memorize all 4,353 gadzillion variations (and counting) so I can win by memorizing all the moves. It’s madness. More pointedly, I doubt if I could do that if I wanted to at this stage of my life. I don’t have the desire to push myself to the upper limits and my age suggests I’m not likely to achieve it.
Okay. Fine. Whatever. I’m sure I’m not alone so, while the game may be doomed for those dudes and dudettes who must reach for the top, for a pawn pusher like myself, not so much. Now here’s where things get interesting.
The Artificial Intelligence Aspect
Ben’s opponent was a computer. More accurately it was a machine with an artificial intelligence engine installed. Now I can tell you (because I imagined it) this AI machine’s capabilities reached far beyond merely playing chess. It was incredibly powerful. However, as the story was likely to progress, it did not revolve around the computer and/or its cohorts taking over the world. No, it was simply a computer. But if you think about it… all of this pulls the curtain back on the world of AI.
The “Chess is Dead” theme did not arise because AI has become so advanced that it can now think better than the best Grandmasters. No. Not at all. So why? Levy’s theme arose because these chess engines are supposedly nearing the completion of calculation all possible chess variations. In short, we humans have finally accumulated enough brute force to get that job done. The same principle applies to all AI.
This is something I’ve been saying all along – Artificial Intelligence is more artificial than intelligent. Computers are great at calculating. They can and do crunch numbers with fantastic speed and accuracy but think about this… if you’ve ever watched anything along the lines of “Good Will Hunting”, you’ll see this immediately. (If the movie doesn’t ring a bell, it is about a Boston kid from the wrong side of the tracks who happens to be a math genius able to solve “unsolvable” math problem. ) What’s the big deal? Those problems really do exist and even the most advanced supercomputers have yet to solve them – or at least some of them. Why? Because those problems are less about crunching numbers and more about actual thinking. In other words…
Computers don’t think. Not really. Yes they do “learn”, meaning they compile data and some can draw new conclusions from the data but any computer or really, AI program, can only venture as far as its programming will allow it. In other words, a computer’s ability to “think” is only as good as the skill of the programmer or more likely, programming team. In the end, either you’ll have a program developed by a single (flawed) programming genius or by a (exponentially flawed) committee. Yeah.
The real danger of the AI craze is not that computers will “take over the world” per se but we humans will be deceived into thinking as much. Chew on that for a bit.