Episode Details

Autonomous vehicles are statistically less likely to hit a pedestrian than you are. So why do nearly 75% of people refuse to ride in one?

In this episode, we cut through the sci-fi marketing and crash headlines to examine what's actually happening on our roads right now. We start with the deceptive language selling you a "full self-driving" car that legally requires your hands on the wheel at all times, then break down the SAE's Level 0–5 scale — and why those levels aren't a fixed feature you buy at a dealership but a dynamic, second-by-second relationship between human and machine that shifts mid-drive.

We dig into the great sensor war between Waymo's laser-powered LiDAR arrays and Tesla's vision-only camera approach, why both philosophies have severe blind spots (literally, at dawn and dusk), and what a landmark 2024 Nature Communications study of 37,000+ incidents reveals about where AI drives better than humans — and where it catastrophically fails. Then we confront the harder questions: the trolley problem encoded in software, Georgia Tech research showing detection systems are 5% worse at recognizing darker-skinned pedestrians, 2.9 million U.S. jobs on the chopping block, a legal system that has no idea who to blame when an algorithm kills someone, and the rolling surveillance machine you climb into every time you tap "Start Ride."

The technology is advancing at lightning speed. The laws, ethics, and public trust required to support it are not.