Yeah. No. No way. My son and I were discussing the state of AI/robotics yesterday, as we walked a beautiful trail in the Sequoia National Forest. What prompte

Yeah. No. No way. My son and I were discussing the state of AI/robotics yesterday, as we walked a beautiful trail in the Sequoia National Forest. What prompted the discussion was a simple question: What would it take to build a robot capable of navigating these trails as we do? This would be a robot able to do this in a manner indistinguishable from, say, a ten year old human. No GPS, maps, compass, pre-mapping, lidar, ultrasonic sensors, etc. Just vision, hearing and touch/force sensing at the “skin” and articulations. What do you know? Two things: You are located at the start of the correct trail and there’s a waterfall at the end. Our conclusion was equally short and simple: Today, it is hopelessly impossible to match what a ten year old kid could do on that trail. Maybe in ten years. Maybe. AI today can’t do what a ten year old human, or the young bears we so along the trail, can do instantly and without thinking: Understand. We just don’t know how to approach and encode understanding yet.