
Chess used to be a game where intelligence ruled. It ruled so much that it became a benchmark for artificial intelligence. Then on May 12, 1997, a computer (IBM’s DeepBlue) beat Kasparov, the chess world champion.
From that moment many people acknowledged that computers are better than humans at playing chess but among these most declared that playing chess is not really a proof of intelligence. That reminded me of the fox and the grapes (Aesop) with the fox unable to reach the grapes stating that it was really not interested in the grapes…
Anyhow, from that moment playing chess against a computer lost interest, there was no chance. As bad as it was it just got worse.
With DeepBlue there were a bunch of people involved to teach the computer. They taught it so well that in the end the pupil got better than the teacher, but at least humans played the teacher’s role. Then, in 2017 AlphaZero, a Google Computer, sibling of DeepMind, took upon itself to lear to play chess by learning the rules and then starting to play against itself. In just 24 hours it got so proficient that it was able to beat the best computer chess player (out of 100 games, it lost not a single one). We humans have been taken out of the loop!
Rather than giving up to the gloom of being beaten by computers a chess master, Vladivir Kramnik, decided to exploit AlphaZero to explore variations to the chess rules wondering how an alteration to a rule, or a few of them, would change the game, resulting in different shades. As an example what would happen if castling wouldn’be be allowed (moving in one stroke the tower and the king?
Rather than fighting a kost battle against computers (and AI) why not team up with them to create something that is beyond human capabilities but that requires human capabilities? This is the basic motivation of Kramnik: discover new variant of chess rules that would produce engaging games. Trying this by himself would be an impossibly lengthy endeavour, teaming up with AlphaZero where the computer brings its brute force AND its machine creativity and the human brings his aesthetic sense is a win win proposition.
This cooperation among humans and AI represent a paradigm shift, and acknowledgement of the intelligence of machines and of its difference form ours. It is no longer a game of making machines smarter than us, rather the goal is to harvest machine intelligence to increase our capabilities.
This is the theme of an upcoming event organised by OTM/Media 2000 , a virtual event played on the cyberspace, addressing the topic of AI at the core of the Digital Transformation. It will run on September 29th and you can join at this link.
This new paradigm will also be addressed, I guess, in the planned special issue of IEEE Computer magazine, edited by Phillip A. Laplante, and Dejan Milojicic, looking into the future: “Crystal Ball for Engineers“. The issue will be addressing a comprehensive slate of future paradigm changes in several sectors, you may want to submit an article to share your thoughts at this link.