"Artificial intelligence should no longer be defined as an opponent to compete against, but as a tool that helps us unleash greater creativity," said Lee Sedol, 9-dan professional Go player.
Lee made the remarks at the "Agentic AI Commercialization Global Campaign" hosted by AI startup Enhance on May 9. The event was held at the Four Seasons Hotel in Seoul—the same venue where Lee's historic match against DeepMind's AlphaGo took place in 2016. The site of the head-to-head battle between human and AI nine years ago has now transformed into a stage for human-AI collaboration.
The highlight of the event was Lee personally planning and building a Go model using Enhance's AI operating system (OS). Lee created the model entirely through voice commands without writing a single line of code. The Go model he envisioned was an AI-based educational program with tutoring features that would allow beginners to learn the game easily. "Programs that play Go well already exist, but there is virtually no educational AI that teaches Go," Lee said. "If AI can teach Go, the barrier to entry for the game will be significantly lowered."
Based on conversations with Lee about the Go model, AI agents generated meeting notes and project proposals. They then conducted research on Go and education, and began coding work reflecting Lee's preferred design. Multiple agents handling secretarial tasks, design, and program development were deployed. The entire process took only about 20 minutes. "I've tried vibe coding before, but it wasn't easy when I knew nothing about coding," Lee said. "This time, I barely did anything, yet the model was completed." He was visibly amazed.
The quality of the finished program was also impressive. After testing it, Lee said it "surpassed the level of AlphaGo," adding that "the speed and disruptive power of change is faster than I expected."
These capabilities were made possible through Enhance's integrated system called "AI OS," where multiple agents handling web search, shopping, planning, design, and coding collaborate simultaneously. When a user makes a voice request, the AI OS recognizes it and distributes relevant tasks to agents with specific roles. This reduces the time and cost required for program development.
Anthropic's Claude was used for the AI OS's coding tasks. Anthropic has disrupted the software-as-a-service model with products such as Claude Code and Claude Cowork. Anthropic, along with Nvidia and Microsoft, participated as official sponsors of the event. Enhance plans to focus on selling its AI OS to global companies following this campaign.
"If the AlphaGo match symbolized competition between humans and AI, AI is now becoming a collaborative partner that solves problems alongside humans," said Lee Seung-hyun, CEO of Enhance. "We will standardize Enhance's core technologies—ontology, agentic AI, and large language models—to usher in an era of AI OS where companies worldwide can collaborate with agentic AI."