Self-Driving Cars Might Not Be Our Future After All
Is it time to reassess our self-driving future?
In response to a tweet this week that praised Tesla’s Full Self-Driving Beta, even Elon Musk couldn’t resist voicing his doubts about the model, replying, “FSD Beta 9.2 is actually not great imo, but Autopilot/AI team is rallying to improve as fast as possible.” It’s not exactly a vote of confidence for autonomous technology, especially not when it comes from a presumed leader in the race to make autonomous cars viable.
To be fair, Musk walked back his statement almost immediately, following up with a breezy, optimistic tweet about the “much improved” FSD Beta 9.3. But at this point in the game, is anyone really benefitting from the rose-colored glasses that Musk and others seem to be wearing? Many of them might be forced to admit defeat after multiple car and rideshare companies have either pushed back their self-driving deadlines or abandoned the effort altogether. The reality is we could be looking at a future where self-driving vehicles never fully materialize – or at least, not in the way we expect them to.
A few years back, when it looked as though artificial intelligence was just about ready to take over our steering wheels, the ability to eliminate human error on the road seemed downright tantalizing. After all, who wouldn’t want a world without reckless drivers, road rage, or preventable crashes? But we failed to consider that, in many situations, it’s the audacity and adaptability of human drivers that keeps traffic flowing smoothly. Variables like inclement weather, jaywalking pedestrians, and roadside debris can all be simulated, but it’s still difficult to teach a machine how to adjust for those obstacles in the moment. So even though human drivers sometimes make stupid decisions, our ability to make spontaneous decisions at all gives us an advantage over AI.
And with plenty of drivers feeling less than enthusiastic about trading in their stick shift for a robot chauffeur, there are also serious doubts about whether self-driving cars can successfully share the road with other vehicles. The tech company Waymo, another leader in the self-driving field, reported 18 traffic incidents involving their Phoenix-based test vehicles in 2020, many of which occurred because human drivers had trouble adjusting to the driving habits of the Waymo vehicles. So, while it’s certainly not realistic to just wait for all cars to be self-driving, the alternative is creating potentially dangerous situations on the road.
Understandably, some companies have already chosen to step back from what was once an enthusiastic commitment to autonomous technology. Both Uber and Lyft have sold their self-driving car divisions over the course of the last year, although that doesn’t necessarily mean they’ve given up on offering autonomous rides – they’re just leaving it to someone else to hammer out the details. But what happens if it turns out that nobody is up to the task? What do we do if the minor tweaks needed to improve this technology turn out to be major problems that take years to solve?
Silicon Valley hopefuls like Tesla may have to focus on semi-autonomous features instead, like their Autopilot advanced driver assist system (which is currently under government investigation for some serious safety issues of its own). It’s not overly pessimistic to wonder if the future of self-driving will look less like driverless taxis and more like upgraded cruise control because right now, we’re simply not getting where we need to be.