Self-driving cars aren’t living up to their promises. For more than a decade, tech companies have insisted they’re the next big thing taking over the roads. So… when will that finally happen?

Yes, the industry is growing. Just look at Waymo. It just announced expansion to more cities, and five of them will have fully autonomous robotaxis. That’s real progress. But it’s still nowhere near the sweeping takeover we were told to expect when Pittsburgh was crowned the birthplace of autonomous vehicles a decade ago.

Car crashes kill around 40,000 people per year. AV companies say they can change that.

Car crashes kill around 40,000 people per year, a leading cause of death in the US. AV companies say they can change that. Robots, they argue, remove the human unpredictability, like sneezing or reading a text or even DUIs. It’s a noble goal with promising data behind it. 

Yet, after all these years of hearing the pitch, the technology still feels stuck in its awkward adolescent phase.

That’s what reporter Alice Crow and I were talking about after the recent Waymo expansion news, including a full robotaxi rollout in Philly, plus testing in Baltimore and Pittsburgh. Self-driving services keep landing in new places, but high-profile errors that get publicized in the news inevitably overshadow the progress, keeping the whole industry locked in this perpetual “almost there” zone.

Part of the problem is structural: regulatory hurdles, expensive innovation, all the usual suspects. 

But part of it is reputational. Most of us like to think we’re better drivers than we actually are, which makes it pretty easy to assume we’re better than robots, too. And at least among the Philadelphians we talked to, there’s a real wariness about a tech overhaul on the roads.

Meanwhile, the safety data keeps stacking up in autonomous vehicles’ (AVs) favor — and not just the intrinsically-biased internal reports. (It’s worth noting that while self-driving cars are closely monitored, we don’t have as reliable data on traditional car crashes, since we have to rely on police reports.) A recent analysis of thousands of car accidents from researchers at the University of Central Florida found that AVs are safer than human drivers “in most accident scenarios.” AVs perform statistically better in ideal conditions, while humans still do better at night, in the rain or when the unexpected occurs.

Waymo is arguably the frontrunner in shifting that narrative. It has the most cars on the road and an uncanny ability to spin almost anything into a positive. 

Even a viral video of one of its vehicles driving straight into an active police standoff became, in Waymo-speak, another learning opportunity: 

“When we encounter unusual events like this one, we learn from them as we continue improving road safety and operating in dynamic cities,” a Waymo spokesperson told ABC7. Tesla, for its part, lost that reputational battle years ago, because of fires, fatal crashes and, well, its CEO.

So yes, there’s progress. But for a lot of people, this so-called miracle tech still feels abstract. It’s a future we’re constantly told is arriving, but it trips over its own reputation before it pulls up.