Recently, the three major turning points in the AI industry discussed by Jensen Huang at the Davos Conference have attracted attention. Personally, what I felt listening to him is that these points quite accurately capture the direction of AI's evolution.

The first transition Jensen Huang mentioned is the emergence of Agentic, meaning autonomous AI agents that have made the implementation in research fields a reality. Until now, models only performed given tasks, but Agentic is beginning to have the ability to think and act on its own.

The second significant development is the rapid maturity of open-source models. Especially with the advent of open-source inference models like DeepSeek, the entire ecosystem has been revitalized all at once. It demonstrated the possibility that areas once thought to be monopolized by major corporations could actually develop in an open manner. Jensen Huang also recognizes this trend as a major industry turning point.

The third development is physical AI, the emergence of AI that understands not only language but the physical world itself. As models begin to understand phenomena such as proteins, chemical reactions, and physical laws, the scope of AI applications has expanded dramatically. From Huang's points, it’s clear that the era of mere text generation is over, shifting toward a deeper understanding of the world.

Looking at these three trends, we can see where the entire AI industry is heading. Jensen Huang’s perspective is quite helpful in understanding the industry’s direction.