Self-Driving Cars Are Miracle Drugs (www.derekthompson.org)

A provocative new piece argues that commercial robotaxis—exemplified by Waymo—are effectively a public-health intervention with the scale to save tens of thousands of lives. Analyses of tens of millions of Waymo miles report dramatic reductions versus human drivers: ~96% fewer vehicle-to-vehicle intersection crashes, ~90% fewer bodily‑injury claims, ~92% fewer pedestrian injuries, and ~88% fewer property‑damage claims. Given that roughly 40,000 Americans die in traffic each year (comparable to major diseases), widespread deployment could meaningfully improve U.S. life expectancy. Independent reviewers have generally endorsed the methods behind these studies, and closer looks at Waymo incidents show many crashes involved stationary vehicles, being rear‑ended, or passenger door mishaps—not core perception or planning failures. The story also flags the political and operational headwinds that could stymie benefits: several progressive U.S. cities are considering bans or rules requiring onboard human safety operators, and deployment is already uneven (common in Phoenix and San Francisco, absent in many other metros). Technical caveats remain—scaling into highways, snow, and dense urban contexts, plus infrastructure, parking and labor‑displacement questions—but the empirical safety delta versus human driving makes the technology a high‑leverage target for policy debate. The debate is therefore less about whether autonomy can improve safety and more about how, where, and under what rules we let it do so.