黄仁勋在G20与美国商务部长Howard Lutnick同台，解释AI的智能体harness。他表示大语言模型（LLM）本身只是大脑，真正让AI变得有用的是在LLM外加上一层"外骨骼"，即agent harness，为大脑提供检索知识、工作记忆、使用工具和协作等能力，从而解决问题。
Jensen Huang explains AI's agentic harness at the G20, with U.S. Commerce Secretary Howard Lutnick .
"AI is intelligence, and it starts with software.
In fact, if you look at AI in the last year, we started by talking about large language models, or LLMs.
So, large language models, what made AI useful was putting an exoskeleton around the LLM.
That exoskeleton is what is called an agent harness.
It essentially gives the brain the parts necessary to retrieve knowledge, have working memory, use tools, collaborate, and so on, in order to solve problems."
----
From "DRM News" YouTube channel, (full video link in comment)
来源：@rohanpaul_ai · x.com