It has been nearly a decade since Google DeepMind’s AlphaGo stunned the world by defeating legendary player Lee Sedol, and the game hasn’t just moved on: it has fundamentally mutated. A fascinating report from MIT Tech Review reveals that top-tier Go players aren’t just losing to AI anymore; they are actively rewiring their brains to think like machines. The days of human intuition and “creative” style are fading, replaced by a rigorous adherence to algorithmic perfection.
The Rise of “Shintelligence”
The report highlights the routine of Shin Jin-seo, the world’s top-ranked Go player. Shin doesn’t just practice; he engages in what he calls an “ascetic practice” with an AI program called KataGo. Every morning, he studies the AI’s suggestions, specifically the “blue spot” on the screen that indicates the statistically optimal move.
Shin’s approach has earned him the nickname “Shintelligence” because his gameplay mimics the machine so closely. According to data cited by MIT Tech Review from the Korean Baduk League:
- The Average Pro: Matches AI suggestions about 28.5% of the time.
- Shin Jin-seo: Matches AI suggestions 37.5% of the time.
This gap is massive in a game of such complexity. Shin admits that he often plays moves he doesn’t fully understand simply because the AI suggests them. He spends his time reverse-engineering the machine’s logic rather than developing his own human theories.
The Death of Human Intuition?
This shift raises a massive question for the industry: Is AI killing creativity? The article notes that players used to invent their own styles. Now, playing “professionally” means memorizing and replicating AI patterns. If you deviate from the AI’s path to try something “creative,” you will likely lose to someone who stuck to the script.
However, there is a silver lining. I found the report’s insight on democratization particularly uplifting. Because high-level AI training tools are now accessible to anyone with a computer (not just those in elite training schools), the playing field is leveling. MIT Tech Review reports that more female players are climbing the ranks than ever before, directly attributing this success to AI-assisted training.
The “Tabula Rasa” Lesson
For those of us watching the broader AI industry, the evolution of the AI models themselves is the critical takeaway here. The original AlphaGo was trained on 30 million human moves. It was incredible, but it was still tethered to human logic.
Its successor, AlphaGo Zero, learned from scratch: tabula rasa. It played against itself, unconstrained by human biases or history. The result? It crushed the original AlphaGo 100 games to zero after just three days of training. This proves a powerful point for AI development: sometimes, human data is a bottleneck, not a fuel source.
What This Means for You
While few of us are professional Go players, this trajectory mirrors what we are seeing in coding, design, and strategy.
- Augmentation is Mandatory: Just as it is impossible to compete in Go without AI training, professional viability in many fields now requires AI fluency.
- Trust the “Blue Spot”: We are entering an era where the optimal solution might be counter-intuitive. Professionals need to learn when to trust the model over their gut.
- The Human Edge: Shin believes he could beat the original AlphaGo today because he has trained against superior models. Humans are getting smarter because of the AI. The goal isn’t to be the machine, but to use the machine to surpass previous human limits.
Shin is arguably the strongest human Go player in history, specifically because he surrendered his ego to the algorithm. That is a lesson every industry leader should take to heart.
I highly recommend reading the full analysis from MIT Tech Review to understand the depth of this transformation.