Marc Haine explores agentic AI, systems that act as personal agents handling tasks for users. He cites a claim that fully autonomous vehicles could prevent 275,000 accidents annually, but notes human behavior remains the biggest safety variable. He connects this to Turing's test for machine intelligence and contrasts agentic AI with creative AI.
Agentic AI is where it is acting as your agent. You will see this on websites and support portals on the web that you'll type in a problem and AI will try to answer all your problems as you're typing it all in. And so, it's representing you. This is agentic AI is also what would be fueling autonomous vehicles. Picture the video of the tractor. The next evolution of that is that you're not going to have a driver in the tractor. Once you program the computer, the tractor at some point will just go up and down the field on its own doing the spraying because it knows what the rules are. And that's really the difference between creative AI and the agent one. They said if every car on the street was artificial intelligence, was autonomous vehicles, that we would have 275,000 less accidents every single year. When I talked to people who are working on this technology, they said the biggest variable on the streets right now for autonomous vehicles is people. And it was interesting. It's like, how do you know the computer's intelligent? Even voicemail, you'll get a call and the person on the phone is actually not a person on the phone. It's actually AI. This gentleman, Alan Turing, who was the gentleman who created the machine that was able to decode the German messages back in World War II, he said that a computer would deserve to be called intelligent if it could deceive humans into believing that it was human.
