The Un-Dumbing of Chess
Are we getting smarter or dumber?
In the last century or so, there are many skills that have been made available to the vast majority. There have been techniques, tools, frameworks, and more than those, there have been heuristics, rules-of-them, systematic steps, and elaborate theories developed in all branches of learning - from music to creative arts to intense analytical subjects - which makes it easy for everyone to learn and master a skill in simple, easy-to-follow steps.
However, if you look closely, it is true that more and more people are learning skills, but it comes at a cost of dumbing down the skill to a degree, that while we might have a perception and an illusion of knowing a skill or even being good or world-class at it, but we have almost no "intuitive" understanding of the skill.
We know a skill, but as an "external", foreign entity.
Chess is a victim of that too ...
In the earliest days, Chess was a game of intellectual brilliance. Players like Paul Morphy played moves that would shock the hell out of an opponent. But these were instances of purely personal brilliance. This brilliance could not be duplicated or learn.
It was not possible to create another player of the shockingly intuitive display like Paul Morphy.
And then came, a number of chess giants who could see certain patterns and rules which could be emulated by others and play at a very high level of excellence.
They distilled the essence of their insights into simple concepts that could be understood by all, and in the process, developed rules, heuristics, and insights that now form the core of "learning chess".
"Playing by the Book" became not an option, but the ONLY way to play chess.
In the process, from being a game of logical analysis and brilliant flashes of insight, chess became a game of memorizing positions, patterns and moves, following the "BOOK", and playing by the right heuristics.
In short, from being a game of fantastic display of brilliance - for the majority of the chess-players, chess became a total mental grind. Space advantage, Time advantage, Material advantage became such unassailable Laws of Chess that they could as well be included in every nation's Penal Codes, and chess players could as well be jailed or hanged for breaking these laws.
The Dumbing Down of Chess
How much we had dumbed down Chess started to become evident when the Machines got better of the Human chess players. While the initial Machines that became super-successful were bad enough, there came the "Zero" series of machines - e.g. Alpha Zero and Leela - systems which learnt not from other players' games, but from SCRATCH.
Alpha Zero and Leela are way beyond any human chess GrandMaster. Not only can grandmasters not even imagine to win against these systems, they cannot even comprehend what these machines are "thinking" (if that can be called so).
If you watch a video of a GrandMaster analyzing a game by Alpha Zero or Leela, you would always see the emotion - "hey, what are you doing, don't break that rule please..." written all over the video and the GrandMaster's face.
No, They Are Not Breaking Rules
What does "rules of a game" exactly mean?
Rules of a game are a part of a shared consensus that all great players all over the world and across a few generations have come to accept as the "right" way to play.
They seem to be right, because they align with common sense.
They work well, because everyone is a party to the same dumbness. The one who manoeuvres these dumb rules the best excel in the sport.
No, what Alpha Zero and Leela are doing should not be called "breaking the rules, yet winning". They are winning in such a demolitionary way because they are doing what they are supposed to do - as a result of their "from the scratch learning", they understand the intricacies of the position in and out, and they are very well aware that even if they go 4 pieces down, they can still win 20 moves later because of the way they have managed to create an unassailable position for themselves.
It is us who have dumbed ourselves to the extent that we cannot comprehend them at all.
Towards Un-Dumbing Chess
This is a popular puzzle, that was supposedly given to Bobby Fischer twice in his life-time, with the challenge to solve it in 30 minutes. Bobby Fischer failed on both occasions.
In the above position, White is supposed to play and win in 3 moves.
We will use this example to demonstrate a way to restoring Chess back to its level of intellectual brilliance, genius displays and a game of "logical analysis and rigour".
Different players would approach it differently. Let us go straight into a discussion on how to "ReInvent" Chess.
ReInventing Chess
Chess is a game of logical analysis and an almost mathematical and algorithmic rigour.
By nature, an approach to Chess should involve flairs of Analytical Problem Solving as well as Creative Problem Solving.
Here, we would explore, how an approach that is a mix of Analytical Problem Solving and Creative Problem Solving could lead us to a framework of Chess, which would make human beings capable of understanding Alpha Zero and Leela, and in due course of time, surpass them.
ReInvent's 3-Step DIR Framework
ReInvent's 3-Step Framework for Learning, Solving Problems and almost any Cognitive activity involves the steps of :
D - Deconstruction - Breakdown a problem, Get the right representation, Ask the Right Questions
I - Insight - If the Deconstruction is right, it will lead to an "Insight"
R - Reconstruction - Once the "Insight" is in place, the problem is considered solved. We need to work further to complete the details of the solution.
Deconstructing the Bobby Fischer Puzzle
Let us look at how to deconstruct the Bobby Fischer Puzzle.
The White has to play and mate in 3 moves.
When we try to deconstruct this problem-statement, we might ask -
"What would be the final mating position look like?"
To be able to answer this question appropriately, we need to deconstruct even further, and ask the most fundamental question of Chess -
"What is a checkmate?"
Yes. That's where we will start this discussion. This discussion starts by trying to answer "What is a Check-Mate?"
A check-mate is a situation in which the
(1) The King is attacked. This is a normal "check". If the King can move elsewhere, it can be saved. A checkmate comes, when additionally,
(2) The King cannot move to an unattacked square.
This could even happen if the King is at the edge already and cannot move further.
The bottom-line is - the King is attacked, and it cannot escape it.
领英推荐
CheckMate in the Bobby Fischer's Puzzle
So, what are the possible check-mate positions in Fischer's Puzzle?
It seems intuitive that the Killing attack will come from the Queen, and that the Queen should be just next to the King. (More detailing than this will require a lot of discussion and we will avoid that in this article. But if you think a while, you will see that this is quite obvious).
Now, what squares, does this Queen attack?
The Queen attacks the King and almost every other square adjacent to the King, except for two "escape holes".
These two squares can be attacked by a Bishop.
What does this say about the role of one of the two Bishops in the puzzle?
So, we have a position which is indeed a check-mate, requiring one additional condition - the Queen needs to be protected too.
Who will protect the Queen?
What does this say about the role of the 2nd Bishop?
It can be easily shown that a position when the Queen is to the side of the King leaves two unattacked squares that cannot be attacked by a Bishop, and is not a feasible check-mate position.
Also, it is easy to see that the Queen on any of the diagonal neighbouring squares of King would be a situation worth considering, and in each case there are three aspects -
(1) The Queen attacks the King from a diagonal neighbouring square
(2) One of the Bishops attacks the 2 "escape-squares" that are not attacked by the Queen
(3) The Other Bishop protects the Queen.
What did the "Deconstruction" achieve?
The Deconstruction process revealed to us two important insights -
(1) The potential final positions of the Check-Mate
(2) Role of each of the 3 White position in the Check-Mate
This leads to a lot of other quick corollaries -
(1) Since the Queen delivers the killing blow, the 3rd (final) move of the puzzle would be a Queen move, to a position next to the King
(2) Two other moves remain, and the purpose of those will be to curtail the movement of the King, so that the King could be forced to move to a position with the above conditions satisfied.
Insight - Don't we know EXACTLY what lines to think on?
We are still far from a solution, but we have got an appropriate "problem statement". Getting the problem-statement is the most important step of problem-solving.
A Problem-Statement is very different from the initial Problem Description. Problem-Statement, is a very rigorous and creative step, that helps us come with a very tangible, structured objective to an undefined, random, chaotic, problem description.
We have not only got the problem-statement, but the problem-statement directly leads to an "insight" of what the nature of the solution is going to be.
Reconstruction - Filling in Details
Now that we know exactly what we should aspire for towards a solution, we still need to fill in the details of finding the exact moves for the first two moves of the solution. However, we know that these two moves could probably involve the two Bishops, and might require us to throttle the movement of King and force it to a desired square that leads to one of the positions discussed above.
What did this really mean?
This is not about this particular chess puzzle. This is about a "paradigm shift", a new way at looking at chess, modelling chess from its first principles.
Most players try to solve this problem in a forward way - find the first move, the second move, and the third move. The size of the possible values explode and they give up.
Well, the size of possible values explode so much that even Bobby Fischer gave up!!!
However, first setting a clear-cut objective for the position gives you a "reason" to make your moves. The difference between "making a move and seeing what happens" and "making a move with a clear-cut objective in mind" is PHENOMENAL.
Many players who understand the above analysis well, and try to find the first and the second moves with the clear-cut objectives detailed above, are able to solve the problem within minutes. The same players, just thinking forward, do not reach anywhere.
The article shows, how, by using a method of Deconstruction, you first get towards setting an objective, setup a problem-statement for THAT situation. Each board position will be unique, but the deconstruction will give you a totally different perspective of the game as opposed to what a consideration of "space, time and material advantage" would give.
This is about developing the future of Chess, the future of Artificial Intelligence, that is based on "concepts" rather than heuristics and "an evaluation of all possible positions".
Going beyond "The Rules"
We know that there is more to playing chess than to gain space, time and material advantage. A chess game is eventually all about the "position". I may have all my pieces on the board, and may get check-mated by the opponent who only has a Queen and another piece remaining on the board.
The rules of Chess-playing blind us from being aware of the EXACT intricacies of the position
Un-Dumbing All Disciplines
What is true for Chess is true for all disciplines. In an endeavour to making a skill available to EVERYONE, we are dumbing down every skill to a level where we are losing touch with the real intricacies of the skill.
This is leading to severe generational intellectual decadence. More and more individuals are associated with the skill, but the intellectual depth, the flashes of brilliance, the genius insights, are taking a big blow.
Rather than dumbing down the skill to suit everyone, we need to find ways to intellectualize individuals upwards.
And there is a way out... That's we are upto...
ReInvent is on its mission to reinvent the way we look at world. Be a part of this "ReInvention Revolution".
PS - A description of the complete solution to the puzzle from my favourite Chess YouTube channel - https://www.youtube.com/watch?v=tvMftUQJl5Q
PPS - Anyone interested in discussing this puzzle further, or to know more about ReInvent's DIR Framework may please drop me a note.
(c) ReInvent Software Solutions. All Rights Reserved. 2021.
Founder and Director at R. R. Logic Systems Private Limited | Consultant - Compilers at Helprack
3 年Navin, the primary premise of this article reminds me of a quote of Einstein that I often quote in such contexts: "Things should be made as simple as possible, but no simpler". The statement is extemely profound, especially the second part of it. Unfortunately, the profundity of this statement is lost to most. And that is the root cause of the phenomenon that you have assailed in this article. Can't agree more with you!
Innovator ? Solutions Provider for Embedded Systems and Smart IoT devices ? Engineer ? Precautious AI+IoT Advocate
3 年hmmm...... AlphaGo.
Neophile Researcher| Diverse Technologist|Engineering Generalist| AI Pragmatist | Business, Competitive Solution Analyst | Creativity, Innovation, Meta Cognition Coach |Yoga Philosophy Student |People, Process Enthusiast
3 年Very Nice. Especially the games like Chess have a possibility space and a delayed reward, which makes it more complex for human internalization. This also has been a challenge of Reinforcement Learning Algorithms. Its another thing system can know basic rules play so many games, and remember states. Leaving AI aside, You make a excellent call for action, that rules from the past were actually insight discovered by the players. Your highlight is extremely thoughtful to say, we just consume the rule(set) without even knowing about the discovery or experiencing the discovery in oneself. While its good to know a rule(set) the same knowing almost becomes an inhibition of any new insights emerging from oneself. With this backdrop, you make a leap into an thoughtful approach of ReInvent's 3-Step Framework for Learning, Solving Problems and almost any Cognitive activity involves the steps of : D - Deconstruction - Breakdown a problem, Get the right representation, Ask the Right Questions I - Insight - If the Deconstruction is right, it will lead to an "Insight" R - Reconstruction - Once the "Insight" is in place, the problem is considered solved. We need to work further to complete the details of the solution. and walk us through with an example, and focus us more on path of exploration. Thanks for your detailed blog.
Staff Engineer - Product and Platform Engineering at Altimetrik | Intuit
3 年Very insightful !!