Regulator / Policymaker

For decades the American auto industry held that crashes came from "the nut behind the wheel," and that a safe car was not possible. Volvo made the three-point belt standard in 1959 and released the patent. Detroit watched and did not move.

Regulation moved it. Drugs followed the same path: proof of safety before market arrived in 1962, after thalidomide, not before.

Industry blames the user, regulators act, safety improves, and within a generation the intervention looks obvious. Software is still on the first step.

Discussion prompts

• Detroit's position was that a safe car was impossible. What is the equivalent claim in software today?
• What is the strongest argument against regulating software safety, and how much of it is right?
• Both of these interventions followed a catastrophe. What would software's look like, and can we act without one?