Lesson 17A: Modern Chess AI Whitepaper

Shatranj.ai curriculum publication · Modern chess AI · AI literacy · Educational technology

This whitepaper supports Lesson 17A of the Shatranj.ai curriculum and introduces students, teachers, and youth workers to the major ideas behind modern chess artificial intelligence. It explains how chess engines evolved from classical search-based systems toward modern machine-learning approaches, helping readers understand why chess became one of the most important testbeds in the history of artificial intelligence.

The paper connects practical chess concepts with computer science ideas such as position evaluation, game trees, search depth, heuristics, neural networks, reinforcement learning, and self-play. Rather than treating chess AI as a purely technical subject, it frames chess engines as learning tools that can help young people understand how machines compare possible futures, evaluate decisions, and improve through feedback.

Within the Shatranj.ai project, this publication helps bridge historical chess culture and contemporary AI education. It supports classroom discussion, coding activities, and project-based learning for students who are beginning to move from playing games to understanding how intelligent systems are designed.

Read `Lesson 17A: Modern Chess AI Whitepaper` PDF

Original file: Lesson_17A_Modern_Chess_AI_Whitepaper_v14.pdf
Topics: modern chess AI, chess engines, AlphaZero, Stockfish, search algorithms, neural networks, reinforcement learning, AI literacy, chess education, curriculum design.

From Players to AI Architects

Shatranj.ai project whitepaper · Youth AI education · Computational thinking · Board-game-based learning

This whitepaper presents the educational vision behind the Shatranj.ai project: helping young people move from being only players of games to becoming designers, thinkers, and builders of intelligent systems. It argues that chess and historical board games can serve as accessible gateways into artificial intelligence because they naturally involve rules, choices, strategies, patterns, data, prediction, and decision-making.

The publication explains how board games can support computational thinking without requiring students to begin with abstract programming concepts. Students first encounter familiar ideas: legal moves, good moves, bad moves, plans, threats, sacrifices, and consequences. These ideas can then be connected to AI concepts such as algorithms, evaluation functions, search trees, training data, simulation, optimization, and human-machine collaboration.

The paper also emphasizes the broader social and educational mission of Shatranj.ai. Historical games such as shatranj, chess, and related board games carry cultural memory across many regions and civilizations. By combining cultural heritage with AI literacy, the project creates a more inclusive pathway into STEAM learning, especially for young people who may not initially see themselves as programmers, engineers, or AI researchers.

Read `From Players to AI Architects Whitepaper` PDF

Original file: Whitepaper_From_Players_to_AI_Architects.pdf
Topics: AI education, youth learning, computational thinking, chess-based learning, historical board games, STEAM education, algorithmic reasoning, data literacy, responsible AI.

Suli-Karatekin Diamonds: Reverse-Ferz Shatranj Positions Reaching 63 Plies

Author: Tamer Karatekin · Shatranj.ai research publication · Historical chess · Tablebase analysis · Zenodo DOI

This publication documents the Suli-Karatekin Diamonds, a family of reverse-ferz shatranj endgame positions related to the famous Suli’s Diamond problem. Suli’s Diamond is one of the most celebrated historical shatranj studies, traditionally associated with the early chess culture of the Islamic Golden Age and remembered for its extraordinary difficulty.

The paper extends the study of Suli’s Diamond through modern tablebase analysis. Under shatranj rules, the ferz moves one square diagonally and is much weaker than the modern queen. By reversing the placement of the ferzes and systematically analyzing related positions, the study identifies a family of extremely difficult endgames. The hardest verified Suli-Karatekin Diamond positions require 63 plies to win, exceeding both the canonical 39-ply Suli’s Diamond solution and John Tromp’s 53-ply Rough Diamond.

The publication connects historical chess research, endgame composition, mathematical search, dynamic programming, and artificial intelligence. It shows how a puzzle preserved from the deep history of chess can still generate new discoveries when examined with computational methods. In this way, the paper supports the larger Shatranj.ai mission: using historical board games to teach AI, algorithms, data, and cultural heritage together.

Suggested citation: Karatekin, Tamer. Suli-Karatekin Diamonds: Reverse-Ferz 4-piece Shatranj Endgame Studies Reaching 63 Plies. Zenodo. DOI: 10.5281/zenodo.19465731.

Read Suli-Karatekin Diamonds PDF View Zenodo DOI Record

Original file: Suli_Karatekin_Diamonds_Reverse_Ferz_40_positions.pdf
DOI: https://doi.org/10.5281/zenodo.19465731
Topics: Suli-Karatekin Diamonds, Suli’s Diamond solution, As-Suli’s Diamond, Suli’s Diamond, shatranj, reverse-ferz endgame, tablebase analysis, historical chess, chess AI, artificial intelligence, dynamic programming.