The Information has an “exclusive” today about the problems with Waymo’s “big ambitions.” (Sorry. Paywall.) Waymo is the subsidiary of Alphabet (formerly Google) that runs the company’s autonomous vehicle development. The article chronicles the myriad of problems that the cars, currently being tested in Phoenix, are having navigating the city’s traffic.
The Waymo vans have trouble with many unprotected left turns and with merging into heavy traffic in the Phoenix area, especially on highways. Sometimes, the vans don’t understand basic road features, such as metered red and green lights that regulate the pace of cars merging onto freeways. …The biggest issue for Waymo’s vans and other companies’ prototypes is human drivers or pedestrians who fail to observe traffic laws.
The technology for self-driving cars simply hasn’t progressed all that far, despite all the hype that the industry has tried to generate over the last decade or so – all the predictions that, as this headline from Forbes exemplifies – we are just a few years away from roads that are dominated by autonomous vehicles.
I’ve never been particularly bullish about these predictions. And it’s not simply that, as The Information points out, the technical challenges are still quite significant. I think there are major cultural issues, political issues, and (literal) infrastructure issues too that stand in the way of the rapid timeline that futurists like to give.
Personally, I’d prefer to see greater investment in public transportation than in cars, and I’d rather hear stories that predict that sort of future.
I have long argued that predictions about the future of tech are more about marketing an idea – softening us up for it, if you will – than they are about accurately gauging when something is going to come to pass. And as such, it’s important to take note of who is making the predictions and spreading the predictions. Automakers have invested a lot in the autonomous vehicle technology. Tech companies like Google have invested a lot in it. Unsurprisingly, the people who work in or adjacent to tech are often the ones who repeat these stories casually, insisting that self-driving cars are right around the corner.
It’s been almost six years since I rode in one of Google’s self-driving cars. I think about all the data that Google has amassed since then – all the mapping data and geolocation data and sensor data and historical data and traffic data and all the machine learning that their machines are supposedly doing with that. Why, it’s almost as if the problems of navigating the world with AI are much, much harder than engineers imagined.