Autonomous driving, Steve Wozniak, and the danger of inflated expectations

October 16, 2018

I read recently that Apple co-founder Steve Wozniak had lost hope about the promise of AI-powered autonomous vehicles.

Speaking at a conference in Barcelona, ‘Woz’ told the audience he didn’t think artificial intelligence will ever really be up to the task of powering self-driving vehicles.

“Artificial intelligence in cars is trained to spot everything that is normal on the roads, not something abnormal,” he said. “They aren’t going to be able to read the words on signs and know what they mean. I’ve really given up.”

Coming from a man who helped build one of the world’s most successful companies — and a company with a growing fleet of autonomous test vehicles — I couldn’t help but take notice.

Did his comments reveal something about Apple’s own autonomous vehicle strategy? When a visionary like Wozniak “gives up” on an emerging technology, does that mean we should too?

Self-driving hype

While I can only speculate about Apple’s plans for self-driving cars, I know Wozniak is far from the first person to express skepticism about autonomous vehicles.

From outright doomsday prognostications about the “death” of self-driving to Toyota Research Institute executives who “don’t by the hype” about the inevitable and imminent arrival of full autonomy, there are plenty of doubters out there.

But why are these voices starting to emerge now, right when it feels like there is so much momentum and excitement about the promise of autonomous vehicles?

If you ask me, this chorus of sober realists has always existed. And in some ways, I’m one of them.

While I certainly haven’t “given up” on autonomous driving, I’ve long had concerns that all the hype could have unintended, negative consequences that end up hurting drivers and the entire industry.

Autonomous driving realities

No, the movement of autonomous vehicle realists from the fringes to the front page isn’t a sign of the self-driving apocalypse. Rather, it’s simply a signal that autonomous driving is following a trajectory like many other emerging technologies before it.

Having drunk a few many cocktails spiked with turbo-charged tech euphoria, we are now waking up to a painful hangover of doubts, disillusionment, and diminished expectations.

While this descent was perhaps inevitable (and likely hasn’t been helped by the recent troubles of chief self-driving evangelist Elon Musk), it’s hardly cause for alarm.

In fact, it’s healthy and necessary.

Now that the party is over, everyone can now refocus their energy on the massive and incredibly demanding task of pushing the auto industry even further down the road to full autonomy, one careful step at a time.

Eye-tracking and driver awareness

As a first step, we need to get serious about educating consumers not only about the capabilities of autonomous driving systems, but also about their limitations. A recent report from the AAA in the US found that many drivers are woefully unaware of the limitations of various driver assistance technologies.

“There is a still a lot of work to be done in educating drivers about proper use of (advanced driver assistance systems) technologies and their limitations,” says Dr. David Yang, executive director of the AAA Foundation for Traffic Safety.

I couldn’t agree more. And it looks like things are moving in the right direction.

Consumer Reports recently published its first ranking of automated driving systems – a key step toward improving driver understanding. Among other things, the tests looked at what the systems do to keep drivers engaged — an important signal to drivers that they do in fact still need to drive even if there is an increasing amount of whiz-bang tech to help them out.

(And I can’t say I was surprised to see the car that topped the ranking was the only one to feature an eye-tracking.)

Slope of Enlightenment

All of us in the auto tech industry have a responsibility to make the transition from semi-autonomous to fully-autonomous driving as safe and smooth as possible — especially as it becomes apparent this transition is likely to be a decades-long rather than years-long process.

Indeed, the road from SAE Level 2 to Level 5 is a lot longer than many were led to believe during the period of “peak hype”. That translates to hundreds of millions of vehicles with advanced driver assistance systems that nevertheless fall short of full autonomy.

Thankfully, technologies like the eye tracking systems developed by Smart Eye and other AI-powered driver monitoring systems can help recalibrate expectations, keep people safe, and rekindle a more informed enthusiasm for autonomous vehicles.

So, while many autonomous driving hopefuls may be suffering from a collective chill as we rush into Gartner’s “Trough of Disillusionment”, I’m confident the growing ranks of more informed consumers will see autonomous driving climbing up the Slope of Enlightenment before long.

Martin Krantz is founder and CEO of Smart Eye, a leading eye-tracking technology provider based in Gothenburg, Sweden.

Written by: Martin Krantz