Solving Chess and Societal Issues All at Once
This guest post was written by Karel van Delft and Dries Wedda.
Karel van Delft is a psychologist, chess trainer, and the Science Project Manager at Chessable. He is the author of the book Chess for Educators and co-author of the book Developing Chess Talent. He runs Schaakacademie Apeldoorn (Chess Academy Apeldoorn) in the Netherlands.
Dries Wedda studies artificial intelligence at Rijksuniversiteit Groningen. He is an online international chess trainer from the Netherlands.
Photos: Karel van Delft
Solving chess, helping seniors through chess, and being a community centre are focus points for the Max Euwe Centre.
“We want to contribute to society, I want to stimulate and connect people,” says Professor Jaap van den Herik, president of the renowned Dutch institution since June 2024.
In the Solving Chess project, research methods are being developed that can be used for many scientific and societal issues. “In the footsteps of Euwe, these initiatives involve community building, education and knowledge acquisition,” says Van den Herik. “No problem if the game is solved in 2035, which was my prediction in 2005. It will stay a sport for people, but chess is also a stepping stone to solve even more complex problems.”
The Max Euwe Centre (MEC), located in the heart of Amsterdam, was founded in 1986. Named for Dutch World Champion Max Euwe (1901–1981), the MEC has sports, scientific, cultural, educational and social functions. It includes a freely accessible museum about Euwe, a chess library, and is a venue for exhibitions, lectures, training courses and meetings. The centre organises numerous initiatives, such as an annual festival. Information is shared via its own website https://maxeuwe.nl/en/home-english and elsewhere. The centre is a foundation, with around twenty volunteers and part-time paid manager Jeroen van den Berg (who has been tournament director of the Tata Steel Chess Tournament in Wijk aan Zee since 2000).
Solving Chess
An important project of the MEC is research to ‘weakly’ solve chess. This means that from the starting position a certain outcome is forced assuming perfect play from both sides (white wins, black wins or a draw). This in contrast with ‘strongly’ solved, which means from every possible game position you can determine the game-theoretic result.
During research and during play you can develop research techniques that have broad scientific and social significance and more application possibilities. Compare it with the development of the neural network AlphaZero that has led to the analysis of protein structures and medicines with the neural network AlphaFold. It is all a matter of all kinds of pattern recognition and combinatorial issues. It is for this finding that Demis Hassabis received the 2024 Nobel Prize together with John Jumper.
One of Van den Herik’s Ph.D students, Dr. Victor Allis, solved ‘Connect Four’ (Four in a row) in 1988. Van den Herik is full of energy and enthusiasm: “That can be done also on a larger scale, up to 32 pieces. There are already tablebases with all solutions for positions with seven pieces. It is all about knowledge representation, pattern recognition and search methods. Better hardware and new techniques have become available, for example better search algorithms, advancements in deep learning, quantum computing and human-computer collaboration.”
Search techniques can be improved by removing weaknesses, says Van den Herik. “Also we should remove human-made evaluation functions. Human knowledge is worthless. Human opening books are chock-full of errors. Deep learning is the future: let computers play against themselves, register which patterns and features show up.”
Van den Herik is optimistic about the possibilities of the Solving Chess project. As a scientist and in social functions, Van den Herik has always been a connector. “You achieve more if you work well together and are open to the opinions of others. You must also dare to revise your own opinion.”
Helping Seniors through Chess
The Chess for Seniors project will be developed this year. The goal is teaching chess to seniors to slow down or even prevent Alzheimer's or dementia from occurring. The Max Euwe Centre will cooperate with neuroscientists and intends to promote insights about the benefits of playing chess among senior citizens, and government and social organizations. To develop several initiatives there are contacts with grant providers, the centre will consult with other organizations and organize chess courses.
About Jaap van den Herik
Jaap van den Herik (born 1947) is emeritus professor of computer science. He graduated cum laude in mathematics from the Vrije Universiteit Amsterdam. He obtained his doctorate from the TU Delft. As scientist and professor, Van den Herik worked at universities in Amsterdam (VU), Delft, Maastricht, Tilburg and Leiden and contributed to mathematics, computer science and jurisprudence. He has supervised 96 Ph.D students.
His dissertation Computerschaak, schaakwereld en kunstmatige intelligentie (in Dutch, translation: Computer chess, chess world and artificial intelligence) was one of the first Ph.D. theses in the Netherlands on artificial intelligence. With his dissertation he worked further on insights earlier developed by others, e.g. Claude Shannon who was the first to publish on computer chess in 1950 with his famous article Programming a Computer for Playing Chess. One year later, Alan Turing created the first computer chess playing algorithm.
His supervisors were professors Henk Lombaers (programming), Adriaan de Groot (methodology, thinking psychology) and Joop Doorman (science philosophy). Van den Herik knew De Groot from blitz tournaments at the home of Dutch GM Lodewijk Prins. He differed fundamentally with his supervisor De Groot on the question of whether intuition could be programmable, which the author of Thought and Choice in Chess considered was impossible. That was in 1983, in 1997 Deep Blue defeated Garry Kasparov.
“I argued that intuition is based on knowledge stored in the unconscious or subconscious of people and that one of the tasks of Artificial Intelligence was and is to make that knowledge explicit.”
How did Van den Herik know that computers would become stronger than people? “You need computers that search deeper, if you search more deeply you know more, if you search the entire search tree, then you know everything.” Van den Herik based his view on the increasing power of technology: processor technology, enlargement of memory cards and new search techniques (pruning methods to search more specifically by excluding certain variations).
Van den Herik found little support when he stated during his dissertation presentation in 1983 that computers would become stronger than the World Chess Champion. “Only grandmaster Genna Sosonko agreed with me. He told me about his grandmother in Russia who did not understand that a human voice could come from a box. That was a radio. Just because you do not know something does not mean that it cannot exist.”
As a chess player, Jaap van den Herik was active at a high level. In 1968 he won the Dutch Student Championship and in the same year he became club champion of the Netherlands with SV Rotterdam. His best game was a victory over GM Mikhail Tal in a simultaneous exhibition. He won against GM Paul Keres and played a draw against GM Mikhail Botvinnik. “In simuls, of course.”
During his student days he was a tournament organizer at his club and he provided commentary at major tournaments.
He no longer plays chess himself, because his level dropped over the years. In science and chess Van den Herik always has been ambitious. “Making too many bad moves leads to self-hatred.”
In the computer chess world he is not only known as a scientist, but also as an organizer and author.
He co-founded the Dutch Computer Chess Association in 1980 and served as the Editor-in-Chief of the ICGA Journal from 1983 to 2015. He has also been involved in organizing events like the World Computer Chess Championship and the Computer Olympiad.
Van den Herik knew Euwe well. Euwe helped him develop computer chess, although he did not believe that computers would ever be stronger than the world champion. Together they also gave lectures on computer chess at chess clubs. In 1978 Van den Herik was at Botvinnik’s home in Moscow, which Euwe had arranged (“I’ll just give him a call.”) He had already drawn against him in a simultaneous match. After his death, a nephew of Botvinnik made sure that Van den Herik received his documentation on computer chess.
References
https://maxeuwe.nl
https://maxeuwe.nl/en/home-english
Publication list Jaap van den Herik: https://dblp.org/pid/h/HJaapvandenHerik.html
Valedictory lecture, Leiden 2021
In English The Power of the Blind Spot
https://www.student.universiteitleiden.nl/binaries/content/assets/governance-and-global-affairs/cpl/afscheidscollege-van-den-herik-eng-totaal-2.pdf
In Dutch: De Kracht van de Blinde Vlek
https://scholarlypublications.universiteitleiden.nl/access/item%3A3245221/view
Valedictory lecture, Tilburg 2016
Intuition is programmable
https://scholarlypublications.universiteitleiden.nl/access/item%3A2908678/view
Video in the Max Euwe Centre Jaap van den Herik discussing with GM Jan Timman computer chess and endgames studies (in Dutch, but some parts are chess language)
Demis Hassabis – Scaling, Superhuman AIs, AlphaZero atop LLMs, AlphaFold
Jaap van den Herik – Computer chess: From idea to DeepMind
https://journals.sagepub.com/doi/10.3233/ICG-180075
Chess Programming Wiki
https://www.chessprogramming.org/Jaap_van_den_Herik
Interview with Jaap van den Herik, Founding Father of the BNVKI
https://ii.tudelft.nl/bnvki/?p=1790
International Computer Games Association
https://icga.org
ICGA Computers and Games 2024 conference 2024
https://icga.org/?page_id=3907
ICGA General Meeting 2024
https://journals.sagepub.com/doi/full/10.1177/13896911251315003
