No Go? No Problem.

Today’s Marketing Cookie is about embracing challenges that may seem impossible and find success through perseverance and innovation.

I love chess. In the 7th grade, I won a chess competition at the district level. I won again at the northeastern regionals, and then was completely spanked at the national level. It is a game of planning and anticipation where, in every turn, a player has up to 35 possible moves, leading to an astronomical number of potential game paths. The sheer complexity means that simply calculating every possible move isn't enough; you need to teach a computer to “think” strategically. It took a dedicated team of hardware and software engineers at IBM over 10 years to develop Deep Blue, which finally defeated world champion Garry Kasparov, marking a monumental milestone in the realm of artificial intelligence.

Amazing right!? Well, there’s another game that my son and I have recently discovered called “Go.” It is an ancient game from China that has been played without modification for over 2,500 years. It is a beautiful mix of elegance and depth with just black and white stones on a 19x19 grid and only 3 rules. But don’t let the simplicity fool you—there’s a whole world of strategy in every move. Playing Go feels like having a silent conversation, where patience, intuition, and creativity converge. The smooth, tactile feel of placing each stone, the visual contrast of black and white against the bamboo board, all evoke a sense of timeless artistry.

The complexity of Go is truly staggering, with an estimated 10^170 possible board positions. For context, this number far exceeds the estimated number of atoms in the observable universe, which is around 10^80, or one-hundred thousand quadrillion vigintillion atoms. (I didn't make this up.) That's a 1 followed by 83 zeros.

This immense complexity makes brute-force computation impractical, as no computer could calculate all possible moves and outcomes within a reasonable time. The game's deep strategic elements, involving territory control, balance, and influence, add another layer of difficulty to achieving human-level play.

The journey to develop a computer capable of playing Go at a professional level was marked by a staggering number of failed attempts over five decades. Billions were spent on hundreds of early AI programs and research projects, all of which failed and were abandoned. AI methods that worked for chess were completely inadequate for Go due to its virtually infinite number of possible moves. Researchers across the globe continuously refined their approaches, often hitting roadblocks as their programs struggled to understand the game's deep strategic elements.

It was feared impossible, but Demis Hassabis, a chess prodigy and AI researcher, saw the challenge as a matter of "when" rather than "if" it could be done. Despite numerous setbacks and 50 years of failure in previous attempts by others, Hassabis and his team at DeepMind refused to give up. They combined advanced neural networks, reinforcement learning, and innovative algorithms to create AlphaGo. Through relentless perseverance and groundbreaking research, they achieved a historic breakthrough. With over 200 million people watching, AlphaGo stunned the world by defeating Go champion Lee Sedol in 2016. 

This success not only revolutionized the field of artificial intelligence but also demonstrated the incredible potential of human creativity and persistence, leading Google to acquire DeepMind for $500 million. Above all, Demis proved the most daunting problems can be overcome with determination and ingenuity, just as today’s fortune says, “Benefit by doing things that others give up on.”