AlphaGo: The Moment AI Learned Creativity
Definition: An AI system developed by Google DeepMind that plays the board game Go, becoming the first computer program to beat human professionals.
The “Move 37” Moment
In Game 2 against Lee Sedol, AlphaGo played “Move 37″—a move so unconventional that human commentators thought it was a mistake. It turned out to be brilliant. Relevance to Vibe Coding: This proved that AI isn’t just a “search engine” for existing answers. It can find novel solutions that humans haven’t discovered.
Applying AlphaGo Lessons to Code
- Don’t Dismiss the “Weird”: Sometimes the AI suggests a code pattern you’ve never seen. Don’t immediately reject it. It might be a “Move 37″—a more efficient way to solve the problem that you aren’t aware of.
- Exploration vs. Exploitation: AlphaGo won by exploring millions of paths. When you are stuck on a bug, ask the AI to “Brainstorm 5 radically different approaches.” Let it explore.
From Games to Code
DeepMind (creators of AlphaGo) later built AlphaCode, which competes in programming contests. The underlying tech (Reinforcement Learning) is now being used to train the models we use today (like GPT-4’s RLHF).
- Deep Thinking: New models (like OpenAI o1) use “Chain of Thought” similar to AlphaGo’s tree search—they “think” for a long time before writing a single line of code.
Takeaway
Treat the AI as a creative partner, not just an autocomplete. Ask it: “Is there a better way to do this that I haven’t thought of?”