Jensen Huang states that OpenClaw did for agentic systems what ChatGPT did for generative systems.

Jensen Huang states the odds of 100,000 such AI agents building a company like NVIDIA are 0%.

Jensen Huang states that agentic systems can be kept safe by allowing only two out of three capabilities (access sensitive info, execute code, communicate externally) at any time, not all three.

OpenClaw can manage files, set up or cancel meetings, give live work updates, shop and haggle, and make investments autonomously after an initial command.

Jensen Huang states that people in China are currently teaching AI agents to look for jobs, do work, and make money.

Unlike Siri, Google Assistant, or ChatGPT, OpenClaw has persistent memory, recalling conversations and details from weeks ago to improve performance.