With AlphaZero and its Inventor Demis Hassabis, Chess is Celebrating its Own Final Demise

WHO STOLE MY CHESSE?

Outright idolatry and dangerous obsession by humans toward machines is taking on grotesque and obscene proportions.

Is this mania an anxiety-based disorder? or, regrettably, there is something "innate in humans, a strong tendency to worship [AlphaZero, for example, to believe unreservedly in,] to become abject?"

A recent post by a certain Pete, no full name available on his chess.com profile, and yet, the executive editor on the site, titled Updated AlphaZero crushes Stockfish in new 1,000-game match, received the most enthusiastic and overwhelming response from chess.com members discussing exhaustively all the technical stuff, such as the CPT and TPU design, the number of Flops and PoDR.

What is behind this disturbing and obsessive religious zeal by chess players with technology and chess engines like AlphaZero and Stockfish? Here are some questions that come to mind,

1. What do we really need them for?

2. Should we really care so fanatically about and idolize them?

3. Why should we be striving for chess perfection and "objectivity"?

(Imagine another game, like football where every pass, or kick is "scientifically based", perfect, ideal. What kind of game would it be? Who would watch that BS? Inevitably, all game would be tied, something chess is nearing, we saw 12 draws in recent Carlsen Caruana world championship match).

4. And how can all those colorful diagrams and illustrations from Pete's post, together with the CPU, TPU, GPU considerations, number of Flops and PoDR improve human chessplayer’s way of thinking and enhance our enjoyment of the game?

Man vs Machine

.

Again, let's ask that one key question, why we need AlphaZero, and why we should care about it and give it so much thought and attention?

I understand we do need to rely on (software) tools, fast calculation, and precision in many places, like when NASA module is landing on Mars, or when bug-free software helping cardiac surgeons is paramount.

Here's Google DeepMind, a British artificial intelligence company, founder Demis Hassabis,  "I think about AI as a very powerful tool. What I'm most excited about is applying those tools to science and accelerating breakthroughs."

The games of chess and go were the guinea pigs where Demis Hassabis tested the self-learning AI he created. These tools could later be used to solve the bigger, more important challenges in the world, like climate change or drug design. (just to remind Hassabis, learning is not about things, it should be about meaning of things, which his AlphaZero is obviously utterly powerless to grasp; seems to me that sorrowful machine can only memorize patterns and calculate... fast). 

And that's fine, very good, excellent!

But Demis, why did you pick chess? Our game has already been on the ropes since AI, or Artificial "Intelligence" made inroads in chess and AlphaZero's predecessors have made their presence felt for some time. You are rejoneador that finally kills the chess using AlphaZero as a lance of death.

Heidegger wrote The Question Concerning Technology in 1954, long before any hint of digital ubiquity we inhabit today, with chess engines and all, that the problem with technology is that it “enframes” us, preventing us from seeing the world in any other way than the one it offers. Basically, there's no problem with the existence of technology or the forms it takes, but rather our orientation to technology. Heidegger tries to find the "blind spots" in our thinking in our relationship with technology that keep us from a more profound, more "empowering" way of conceiving the world and our place in it.

Carlo Giuseppe Bolmida (Italian), Evoluzione degli scacchisti/Evolution of the chessplayer

.

The chess engines are merely pre-programmed tactical monster crunchers of bits and bytes, and that’s where their abilities stop. They don't understand the game, and there's no even the "S" of strategy in their "knowledge". There's no intuition ("the only real valuable thing," Einstein) either, no imagination, no ability for creative effort, no spontaneity, no fantasy.

(what I'd like to see from AlphaZero and Demis Hassabis is a simple (hah, far from simple for a machine!) comment in plain English to accompany each move, something we see in books, like

"With this surprising pawn sacrifice White opens up a number of lines for his pieces, and obtains a decisive attack, in comparison with which two pawns play hardly any role. It is very important for White, in addition to opening the f-file, to weaken the e6-square, so that Black always has to keep his eye on the various sacrificial possibilities at this point." comment on 11.f5!  from Keres-Fuderer, Gotheburg Interzonal, 1955 in Paul Keres, The Quest for Perfection.

Can AlphaZero or any other of digital creatures do this, present the Ideas and Motives behind the moves, the Meaning? definitely and positively not, and that is the whole point, can AlphaZero improve my (human) thinking? Can they give me Concepts, can they generate for me wise, wonderfully pregnant words and thoughts able to lead and shape my thinking?)

The lack of all these vital ingredients for human thinking we swallow easily as “scientifically based”. And yet, these are the main elements of how humans make decisions in real world, what the Nobel prize winner Daniel Kahneman calls System 1 in his Thinking: Fast and Slow. Most of time in our daily decision-making we use a mode of thinking (without thinking!) that is a reflex, fast, intuitive and automatic. Only when System 1 is unable to find a solution to the problem at hand, System 2, that is slower, more deliberative, more logical and analytical, kicks in. That is the only thing machines excel at, calculation. And this is not a quite correct statement, they only excel at doing it fast.

I'm going to give you an example how it works. There's a story told by the Serbian GM Ljubomir Ljubojevic (from an interview conducted by Evgeny Surov) about intuition and non-intuitive, but "fast" monsters (the reason I used quotation marks will be evident after you read the story). Here's Ljuba,

“Chess engines are not always right. I have experimented with it a number of times, in a position I would make a move using my intuitive judgment and then turn my last generation engine on, to show me that my move wasn’t even the forth, or fifth line of calculation. But after forty hours of constant crunching the machine finally shows that my move was the first line of play. After forty hours, can you imagine that?!”

So, my dear readers, whose thinking is superior and more efficient then?

Of course, chess engines Are faster, and more importantly, they never make a (huge) blunder, that's where they overtake humans who are prone to mistakes, unable to keep up with machine's consistent near-perfect moves ― just one lousy move by a human player and it is over. This perfection on the side of machines actually negates and kills any game ("chess is no more a game," David Bronstein) ― thanks AlphaZero for making chess become a tic-tac-toe game where every move is perfect and there are no... wins.

So, who stole my chesse? (the term as in the 1474 William Caxton's The game and playe of the chesse)

The Chess Machine, Robert Loehr

.

And now back to the question at the beginning. What is it in human nature to unconditionally accepts (digital) idols, and we revoke our own thinking? One possible answer may be this,

"The totalitarian, to me, is the enemy – the one that’s absolute, the one that wants control over the inside of your head, not just your actions and your taxes. And the origins of that are theocratic, obviously. The beginning of that is the idea that there is a supreme leader, or infallible pope, or a chief rabbi, or whatever, who can ventriloquise the divine and tell us what to do. That has secular forms with gurus and dictators, of course, but it’s essentially the same. There have been some thinkers – Orwell is pre- eminent – who understood that, unfortunately, there is innate in humans a strong tendency to worship, to become abject. So we’re not just fighting the dictators. We’re criticizing our fellow humans for trying to short-cut, to make their lives simpler, by surrendering and saying, “[If] you offer me bliss, of course I’m going to give up some of my mental freedom for that.” We say it’s a false bargain: you’ll get nothing. You’re a fool." Never be afraid of stridency: Richard Dawkins’ interview with Christopher Hitchens.

As long as technology is a good servant, not a master, a dictator, that’s fine.

As long as it is a non-dominant partner to our (chess) intellect, not a replacement for it, that’s fine too.

Else, we should abhor it with every bit of ourselves.

Yet, what we can see around is quite a different landscape. Engines actually have already killed chess. At the top level, we see Hybrids play, half humans, half their latest versions of AlphaZero or something. What about the rest of chess population? We have lost ability to think, to analyze our games, to explore creative ideas, flirt and toy with them ― all on our own.

As Kurt Vonnegut put it in his 1952 Player Piano novel, technology first replaced muscle work, then routine work, and now it is real brainwork being displaced... Engines are, lamentably, abusing and devaluing human thinking.

Demis is right, AlphaZero is just a (powerful) tool. As any other tool, it doesn't think, or help you think. It can only help you get the job done, if you may have some ideas and are able to think and juggle them.

Demis' words cited before, "What I'm most excited about is applying those tools to science and accelerating breakthroughs" have thus completely opposite effect. While his research could apply wisely to the benefit of humanity, there is a marked detrimental effect as far as human thinking is concerned. Instead of accelerating, it decelerates it.

Not only that. Actually, with AlphaZero and its inventor Demis Hassabis, chess is celebrating its own final Demise.