A decade after AlphaGo's historic victory, the AI system originally designed to master the ancient board game has fundamentally reshaped the trajectory of artificial intelligence and scientific discovery. DeepMind utilized AlphaGo to address the immense complexity of Go, which features more possible positions than atoms in the observable universe. By combining deep neural networks with advanced search and reinforcement learning, the system learned to play by analyzing expert human games and then refining its skills through hundreds of thousands of self-play iterations. Following this success, DeepMind developed AlphaGo Zero, which learned from scratch without human data, and subsequently AlphaZero. This generalized system taught itself to master various two-player perfect information games, including chess and shogi, in mere hours and discovered novel strategies that surpassed existing specialized programs. These achievements demonstrated that the underlying technology was ready to tackle real-world scientific challenges. The most significant application of this approach has been in solving the protein folding problem, a fifty-year grand challenge in biology. In 2020, the AlphaFold 2 system cracked this issue by predicting the 3D structure of proteins, a capability essential for understanding diseases and drug development. AlphaFold has since folded the structures of over 200 million known proteins, creating a free open-source database used by more than three million researchers worldwide. This work led to John Jumper and the project's lead receiving the Nobel Prize in Chemistry in 2024. The principles pioneered by AlphaGo have since extended into mathematical reasoning, algorithm discovery, and scientific collaboration. Systems like AlphaProof and AlphaGeometry 2 have achieved medal-standard performance at the International Mathematical Olympiad, proving the viability of AI in complex mathematics. Later, the Deep Think mode of the Gemini model achieved gold-medal level results in the 2025 competition. In the realm of algorithms, the coding agent AlphaEvolve discovered new methods for matrix multiplication and is currently being tested for optimization in data centers and quantum computing. Additionally, an AI co-scientist system has validated scientific hypotheses by analyzing decades of literature, matching conclusions reached by human researchers through traditional experimentation. Despite these specialized successes, the ultimate goal remains the creation of artificial general intelligence (AGI). Future systems must integrate world models, such as the multimodal capabilities of Gemini, with the search and planning techniques of AlphaGo. This combination will allow AI to utilize specialized tools, like AlphaFold, while reasoning across different domains to uncover hidden connections. True AGI will also require genuine creativity, moving beyond finding new strategies within existing games to inventing entirely new concepts and domains. Ten years after Move 37, the convergence of these technologies is paving the way for a new golden age of scientific discovery and the realization of general AI.