Supercomputers Are Revolutionizing Chess And Business, But The Human Touch May Still Have The Edge
Last year, I along with 19 other players simultaneously competed against World Chess Champion and Grandmaster, Magnus Carlsen, in front of an audience of several hundred people. His opponents included finalists in a quant competition, where over 30,000 participants were tasked with building complex mathematical models that aim to predict future price
movements of various financial instruments.
Much has been written about how the skills needed to be successful in chess are similar to the skills needed to be successful in business and finance, including problem-solving, pattern recognition and time management, among others. For that reason, these two worlds may
attract similar people, evidenced by a number of high-profile businesspeople who are also USCF-rated chess players, people like Boaz Weinstein (Saba Capital Management), Peter Thiel (Co-Founder of PayPal), David Norwood (Oxford Science Enterprises) and Duncan Suttles
(President of Magnetar Games).
Fewer people are aware of how advances in technology are moving forward the worlds of both chess and quantitative finance along parallel tracks. Just as computers continue to revolutionize the game of chess, advances in data and technology are arming quantitative researchers with more innovative tools to work with.
Supercomputers have given chess players a near-infinite number of possible outcomes and strategies to study. Grandmasters now prepare for major tournaments using computers to learn all the possibilities created by each move, training themselves to try and see into the future of the game. A big upset came to the world of chess, when in 1997, IBM’s supercomputer Deep Blue beat then world champion, Garry Kasparov. The win was mired in controversy because Kasparov accused the machine of cheating. But it was still a watershed moment in terms of man vs. machine. Innovation has moved quickly since that game and top-level chess players now routinely use specially designed software to help them see more lines of play than the human mind can possibly conjure as efficiently.
Similarly, today’s computational power has transformed researchers’ ability to sift through and analyze data and test repeatedly. Processing mathematical models and statistics with large datasets can help to remove the interference of human emotion and biases while uncovering
ideas and predictions that might otherwise be missed. Supercomputers can help enable researchers to process large datasets and complex calculations both quickly and accurately, allowing them to evaluate a large array of assets and market scenarios.
What we’re seeing both in chess and in business begs the question: Are people surplus to requirements when the machines are doing such a good job?
MIT economics professor Richard Bookstaber, stated in his book “The End of Theory” that “No man is better than a machine, and no machine is better than a man with a machine.” Both analytical roles across industries and chess competitors follow that theory. While technological innovation has changed how chess is played at the highest levels and transformed the resources available to researchers, I believe the human touch is still what makes it all work.
Grandmaster Magnus Carlsen has become arguably one of the greatest players of all time, largely by taking what he’s learned from computers and applying a level of creativity that has confounded his opponents. His skill lies in his inventiveness in the face of players who rely on patterns. As Magnus shared with me last year: "Playing chess well now is all about finding ideas that are missed by engines and leveraging our own human perspectives and creativity to find an edge."
Just like an AI chess player, analytical models and their users benefit from the superior processing power of modern supercomputers. And just like an AI chess player, they have their limitations too. Models can falter in the face of irrational human behavior and global unpredictability. Geopolitical upheavals and unexpected economic changes can all confound
computers. I believe an overlay of human insight and intuition are just as important for successful research, no matter the industry, as the models and data they rely on. Ultimately, it is humans who direct the way we use this technology, and so their insight and curiosity is vital to getting the most possible use out of it.
As we look to the futures of chess and business, it’s impossible to know exactly how technological innovation will shape their next evolution. Technology is such a frenetic and fast-moving sphere that we must be as agile as we are proactive. The recent progress in technology and supercomputing are creating new possibilities across industries. As seen at the chess tournament during the International Quant Championship finals, these two domains can attract similar talent due to overlapping skills and interests. Unfortunately, the skills we hone through modelling and analysis will only take us so far in chess. After all, Magnus defeated all 20 of us in just 35 minutes.
Fortunately, many of his opponents’ analytical skills extend far beyond the chess board, allowing them to pursue opportunities that might just be inspired by their passion for the game.
