We’re now several years past the unintended acceleration scandal that cost Toyota billions of dollars over what we now know for sure . But today, our cars are starting to drive themselves. If you don’t see the problem with this, you should.

I bring this up because Toyota’s unintended acceleration mess, and the lessons we learned from it—or more importantly, didn’t learn from it—was the focus of this week’s Malcolm Gladwell podcast, . I was a guest on it to discuss the case, as was Car and Driver’s Eddie Alterman.

It’s very much worth a listen for people who still believe in the importance of human driving.

In case you’ve forgotten, starting in the late 2000s Toyota was deluged with reports of “runaway” vehicles. This led to lawsuits, federal investigations, recalls, , and a black eye it took the automaker years to recover from.

Ultimately, the cause of these unintended acceleration incidents was deemed to be “pedal misapplication”—drivers hitting the gas instead of the brake—or in some cases, sticky floor mats. But reports of flaws in the cars’ electronic control modules were never substantiated. (Toyota ended up to settle those lawsuits, on a criminal investigation, and on recalls anyway.)

Gladwell deems it a “media circus” pushed by embarrassed drivers, regulators, lawyers and other opportunists who didn’t really understand how cars work. If you’re the kind of person who does, it’s a nice bit of vindication:

I’ve used the phrase “car guys” in this episode a few times... what’s interesting about the car guys is none of them doubted ever this is a problem caused by drivers. Because they understand what a car is—. And non-car people have lost sight of that fact.

“The average driver just expects a car to be completely flawless and to save their lives under any circumstances,” C&D’s Alterman adds.

So where are we at now, in 2016? Things aren’t better. Semi-autonomous cars are already on our roads and fully self-driving cars are on the horizon, but we as drivers haven’t shown any progress in becoming safer or smarter.

Maybe Tesla’s Autopilot isn’t truly ready for primetime, but the driver in the fatal Model S wreck this summer was still instead of minding the car, contrary to its instructions. And just this week in China, we saw an Autopilot crash by someone paying even a modicum of attention.

Joshua D. Brown was killed when his Tesla Model S crashed into the side of a tractor trailer on May

The current administrator of the National Highway Traffic Safety Administration, Mark Rosekind, seems to understand that auto safety must be treated as a human problem as well as a technological one. Here’s what he had to say this week, according to :

“In the auto industry, we’re always looking at changing the technology, because changing the human would be really hard,” Rosekind said here today at CAR’s Management Briefing Seminars. “We’re not going to change us. We can change our behavior, but that is really hard.”

Instead, he argued, continually improving safety technology must go hand-in-hand with improving driver behavior to decrease traffic fatalities.

“The technology doesn’t always work, and humans aren’t always perfect, but I think the combination of the two could get us to zero,” Rosekind said.

But we’re not getting better at driving. We’re increasingly trusting technology that is new and unproven that we still don’t understand. The Autopilot crashes prove this. And humans seem destined to increasingly rely on drivers’ aids and self-driving systems when some knowledge about how cars work—and some awareness on the road—would probably go a lot further in making us safer.

Zero traffic fatalities is a noble goal. But even if cars are fully capable of operating themselves, that goal will never be achieved until we can be bothered to take some responsibility for the machines we trust our lives to.