Not long ago, exactly ten years passed since the appearance of the famous 37th move made by AlphaGo, which defeated Lee Sedol. This moment became a true milestone in the world of artificial intelligence.
In March 2016, a Go match took place in Seoul between DeepMind’s development and the legendary world champion Lee Sedol — who was already a 18-time champion at that time. The game ended with a victory for the machine with a score of 4:1, which surprised many with its unexpectedness.
The second game, played on March 10, is especially memorable: in it, AlphaGo, playing with white stones, made a very unusual and unexpected move on the 37th turn. Initially, experts considered it a possible mistake, but this move changed everything: the game turned against the human, and after a few more moves, Lee Sedol resigned — this happened on the 211th move.
This so-called “37th move” forever entered history as a symbol of AI creativity and its ability to beat humans through strategic thinking. It became a vivid example of how a machine can demonstrate unconventional and creative approaches to solving complex problems.
In honor of the tenth anniversary of this event, the DeepMind team and Demis Hassabis published an overview article. It emphasizes that it was precisely AlphaGo and the brilliant “37th move” that demonstrated the potential of such models and inspired scientists to continue developing series of algorithms like AlphaZero, AlphaFold, or AlphaProof. If artificial intelligence is capable of navigating vast spaces of possibilities in Go, it can be successfully applied to solving real scientific problems with similar complexity.
Ten years have passed, and the results have exceeded all expectations:
– A direct descendant of the AlphaGo architecture — the AlphaProof project — along with AlphaGeometry 2, already shows excellent results at the International Mathematical Olympiad, reaching the silver level.
– The AlphaFold 2 model received a Nobel Prize for its ability to predict the structure of nearly all known proteins — about two hundred million!
– And Gemini Deep Think, inspired by the search structure from the AlphaGo architecture, has already won gold at IMO 2025 and successfully solves the most complex mathematical problems of Erdős.
Today, the breakthrough associated with the 37th move continues to influence the development of artificial general intelligence (AGI). We are convinced that AGI will become one of the most important technologies in human history — an instrument to accelerate progress in science, medicine, and industry.
By the way, there is an interesting documentary about AlphaGo. For those who want to take a closer look — now is the perfect time to watch it.
Created with n8n:
https://cutt.ly/n8n
Created with syllaby:
https://cutt.ly/syllaby