No free lunch today
Conceptual abstraction is the base of my method. The transfer of knowledge from one position to another can only be reached by generalisation of the patterns that occur in a position.
That conceptualization requires a lot of work. There is no free lunch here. You can only ask a grandmaster to gather the material and to draft an outline.
So I went to look at AI whether it could help me. I watched a lot of videos and podcasts to see where it is heading. It turns out that AI is suffering from the same problem.
LLM's will run out of input the next year. Everything from the internet is processed, and you cannot scale up the LLM's much further. Hence the ugly begging letters and policy changes at X, Reddit and Meta lately to use your data for feeding their LLM's.
So the contours of the limitations of LLM's start to become clear, Now the hype is nearly over, it is time to ask the question "what's next?". AGI is the next target to pursue (Artificial General Intelligence). The development of AGI has more or less come to a standstill in 2001. The expert systems of the previous century turned out to be too brittle and too rigid, so investments were redirected. Between 2000 and 2010 "big data" was the mantra.
From the 7 areas where improvement was needed to get to AGI, only one area is developed: LLM's. The rest has more or less come to a standstill in the meantime. I's time that investors get over their craze and start to make more sensible decisions.
I digged in what the main problem is with AGI. It turns out that conceptualization is the main problem. The world is too fuzzy and unpredictable for rigid rule based reasoning. So we need more loosely reasoning. Knowledge graphs and neurosymbolic AI seems to be key here.
Recently a new benchmark has been released: ARC-AGI-2025. Ordinary people score about 66% in this benchmark, while the best LLM doesn't exceed 1.5%
So an LLM is not the best tool for helping me with conceptualization. No free lunch today. At least not for a decade, or so.
PART I:
ReplyDeleteWe cycle back to TANSTAAFL: we still have to do the hard work and heavy lifting ourselves. DRATS!
Lately I’ve been carefully examining the concept behind Zugzwang. It was triggered by Erwich’s Pattern 29, given in a comment to the previous blog post.
I brought up the PDF containing the US CHESS FEDERATION OFFICIAL RULES OF CHESS, 7TH EDITION.
[COPILOT INTERACTION]
I asked Copilot:
“Is the term zugzwang used in these rules?“
Copilot’s answer:
‘The term "zugzwang" does not appear in the sections of the US Chess Federation's Official Rules of Chess that I have access to. However, zugzwang is a well-known chess concept where a player is forced to move, but any legal move worsens their position. It often appears in endgames and strategic play.’
‘Would you like me to explain zugzwang in more detail or find examples of it in famous games?’
My directive:
Explain in detail
Copilot’s answer:
‘Zugzwang is a fascinating chess concept that occurs when a player is forced to make a move, but any legal move will result in a disadvantage. The term is German and roughly translates to "compulsion to move." Unlike a regular bad position, zugzwang is unique because the player on move would prefer to do nothing, but since they must play, they end up making their situation worse.’
‘When does zugzwang happen?’
‘Zugzwang most commonly appears in endgames, where players have fewer pieces, and each move significantly impacts the position. It often involves kings and pawns, where one side must move and loses due to an unavoidable setback.’
‘Examples of zugzwang:’
‘King and pawn endgames – If a king is forced to move away from defending a pawn, that pawn can be captured or prevented from queening.’
‘Rook endgames – A player with a rook might have only bad choices: either move and lose their material or allow the opponent to promote a pawn.’
‘Knight vs. king positions – The knight, if forced to move, may lose control over key squares needed to restrict the opposing king.’
‘Zugzwang in famous games:’
‘One of the best-known zugzwang positions occurred in the 1923 game between Friedrich Sämisch and Aron Nimzowitsch. In this game, Nimzowitsch masterfully restricted Sämisch’s pieces until he was forced into a losing move. This showcased the power of positional play leading to zugzwang.’
‘Would you like to see a specific example with a chess diagram, or learn about strategies to avoid zugzwang in your own games?’
[END COPILOT INTERACTION]
PART II:
ReplyDeleteWhat is left unsaid is the reason for that “compulsion to move” (followed by the deterioration in the moving side’s position). The official rules state clearly that the players MUST strictly alternate moves, one move at a time per turn; consideration of the effect of that mandatory rule is not part of the rules. There is no option to just “pass” without making a move when it is your turn to play. Zugzwang is a conceptualization of the implications of this rule and the consequences flowing from it under certain conditions. Usually, the turn to move is helpful, but sometimes it can prove to be disadvantageous or downright fatal. THAT is NOT part of the rules!
Before the chess clock, there used to be a strategy based on Sitzfleisch, a German term that literally translates to "sit flesh" or "sit meat". It refers to the ability to sit still for long periods. The player faced with a Zugzwang situation could simply sit without making a move on the board, hoping that eventually the opponent would grow weary and abandon the game. The chess clock was invented to prevent this unsportsmanlike behavior.
Nothing restricts the concept of zugzwang to endgame positions. That should be obvious, given that the definition of “endgame” is amorphous at best, and the official compulsion to make alternating moves on each turn applies throughout the entire game.
If we conceptualize zugzwang as applicable throughout the game, we gain a valuable tool that expands the range of available forcing moves. Instead of forcing moves being (somewhat) restricted to [CCT] checks, captures (as a sequence to maintain the relative material balance) and threats (either direct attack or threats of multiple orders of distance from a direct attack, always ultimately resolving into a double attack in the broadest sense [Averbakh]), we can now work toward situations in which the opponent’s position deteriorates just a little bit, without necessarily getting to the point that he must experience significant loss (checkmate, serious material loss or positional loss).
In short, we can give the opponent the choice of the lesser of two (or more) evils throughout the entire game, and, because of the compulsory move rule, eventually he must choose one of the not-so-good alternatives. The game Sämisch vs Nimzowitsch illustrates this broadened concept brilliantly in the middle game.
The rules are not always our friend. Most of us unnecessarily limit our understanding and application of zugzwang and similar concepts because we take the underlying official rules for granted.
There is a zugzwang-like mechanism at work. But the two pawns fully compensate for the material loss. So we not only have to study the zugzwang mechanism, but we must investigate the draw mechanisms that remain on the board too.
ReplyDeleteWhat we must realize, is that there are a finite amount of concepts.
ReplyDeleteThey can be combined in an infinite amount of ways. But before you can do that, you must have absorbed them.
In your example there are two concepts that are battling each other. The zugzwang and the draw mechanism against a bishop.
Sometimes you make a multi purpose move, which serves two concepts.
Sometimes you must weigh one concept against another.
Sometimes a concept is met with a counter concept.
It all starts with absorbing the concepts. Only then, the battle of the concepts can begin.
Concepts go beyond rules. An expert system fails, because of rigid application of the rules.