The Intersection of AI and Humanity: How Go Changed Everything
Written on
Chapter 1: The Emergence of AI Through Go
The journey of artificial intelligence becoming a recognized entity began with a game that’s been played for millennia. The ancient game of Go, with its complex strategies, presented a formidable challenge to AI, culminating in a historic clash that would forever alter perceptions of both human and machine intelligence.
“I am confident about the match. I believe human intuition is too advanced for AI to have caught up. I'm going to do my best to protect human intelligence.”
— Lee Sodol, Go Grandmaster (9th Dan) before competing against AlphaGo.
Determining the exact moment an idea embeds itself in society is challenging. Music producer Rick Rubin posits that ideas originate from a universal source, accessible to those with the right sensitivity. In his book, The Creative Act: A Way of Being, he suggests that if an idea isn’t captured, it simply waits for someone who is receptive enough to acknowledge it.
For example, the game of Go has an origin shrouded in mystery, with estimates ranging from four thousand years to two thousand years old. John Fairbairn, an expert on the game, cites its mention in the Analects of Confucius and other ancient texts, indicating its long history. However, the game’s modern popularity surged as it crossed into Japan from Korea, where it acquired the name "Igo."
The term "AI" is now commonly recognized across demographics, similar to how the word "internet" once was. If you ask someone about AI, chances are they will have an opinion or an anecdote to share. This widespread recognition signifies a cultural shift, a tipping point that was irrevocably cemented in March 2016 when computers began to be labeled as "AI" after mastering the game of Go.
Section 1.1: The Complexity of Go
While computers had previously defeated chess champions, Go posed a much greater challenge. In the documentary AlphaGo, director Greg Kohs illustrates why Go is seen as the "holy grail" of computer intelligence. Frank Lantz, the Director of NYU's Game Center, describes the game as “putting your hand on the third rail of the universe,” due to its demanding focus and the way it pushes players to their limits.
The goal of Go is straightforward: dominate the board with a connected group of white or black stones. Yet, the myriad choices available at each turn creates ripple effects throughout the game, mirroring the complexities of life. Demis Hassabis, CEO of Google’s DeepMind, notes that while chess allows for about twenty moves per position, Go offers a staggering two hundred. The configurations on a Go board outnumber the atoms in the universe, making it a seemingly insurmountable challenge for AI.
AlphaGo - The Movie presents an in-depth look at this monumental event, detailing the journey of AI's evolution and its implications for humanity.
Section 1.2: The Rise of AlphaGo
In the documentary, Cade Metz explains that AlphaGo was a multi-stage program. Initially, it utilized deep neural networks to learn from a vast dataset, enabling it to recognize patterns like a human would. This technology is similar to that used in today's Large Language Models, such as ChatGPT. However, AlphaGo incorporated a further level called reinforcement learning, where the program learned independently by playing millions of simulated matches against itself.
The true test for AlphaGo came when it faced the world Go Champion, Lee Sodol.
Chapter 2: The Legendary Showdown
The History of Computer Chess: An AI Perspective delves into the evolution of AI in games and highlights its broader implications on technology and society.
Lee Sodol, a Go prodigy with an impressive record, was anticipated to dominate the match. Yet, the outcome shocked everyone. In a stunning display, AlphaGo won the first two games, leading to a moment of disbelief during the third game when it executed an unprecedented move, dubbed "Move 37." This move baffled Sodol and observers alike, marking a pivotal moment in the history of AI.
The defeat raised profound questions about human uniqueness and the role of machines in creative fields. As Sodol faced the reality of losing to a computer, both spectators and the DeepMind team experienced a sense of sadness, as if a cherished part of humanity was overshadowed by silicon and circuits.
Rick Rubin offers a different perspective on this perceived loss. He argues that AlphaGo triumphed because it approached the game without the biases accumulated over centuries. This "beginner's mind" allowed the AI to explore uncharted territory in strategy, prompting Sodol to adapt and innovate in response.
In the end, AI doesn't have to signify humanity's decline; rather, it can serve as a catalyst for growth and understanding. Just as Go has long been a tool for self-development, perhaps AI can also foster new insights into our capabilities and creativity.
If you wish to explore more about this intriguing intersection, consider subscribing to my mailing list or visiting my profile page for further discussions.