[Jensen Huang Defines a New AI Computing Paradigm: Agent = Large Model + Orchestration Engine, Ushering in a Golden Era for Software Companies] According to monitoring by Beating, Jensen Huang systematically outlined the architecture of the 'AI Agent' during his GTC Taipei 2026 keynote: large language models are responsible for thinking, reasoning, and planning, while external orchestration engines (harnesses) function like operating systems, connecting models to tools such as spreadsheets, browsers, and databases, and managing both working memory and long-term memory. He defined this as a new computing paradigm—shifting from 'launching applications and clicking inputs' to 'describing intentions to AI, which generates code, invokes tools, and delivers results.' Jensen Huang demonstrated several Agent examples live: generating complete application code from natural language prompts; creating a dynamic particle animation transitioning from Taipei 101 → GTC Taipei 2026 → NVIDIA logo based on a brief description; and even taking a photo of a remote control's missing battery clip, with the Agent directly invoking CAD tools to generate a 3D-printable replacement part file. He mentioned, 'Here we’re using Claude Code, but Codex performs equally well.' In response to claims that 'Agents will cause software companies to go out of business,' Jensen Huang once again refuted this: 'Quite the opposite.' He argued that when the world is no longer constrained by the number of humans, a massive number of Agents will use more software tools than humans ever could. 'Now is the best time to be a software company'—but the prerequisite is that software must be presented in a way that Agents can invoke. He also announced that NVIDIA's CUDA X libraries are now fully open to Agents, with Agents using these libraries more efficiently than human developers. [Original Article Link]