It's time to admit self-driving cars aren't going to happen
A couple caveats for those going apoplectic over the headline: I mean self-driving isn't going to be a thing A) in our lifetimes and B) with any kind of omnipresent scale. So in terms of the daily lived experience of most people reading this, truly autonomous vehicles just aren't going to happen. The evidence pointing to this has been mounting for years now, if not decades, but it's now tipped the balance to where it's hard to ignore for a reasoned observer -- even one like myself who has previously been very optimistic about self-driving prospects.
My decision to make this call is mostly predicated on one big event from Wednesday: Scooped by our very own Kirsten Korosec, Ford announced that it would be winding down Argo AI, the company backed by itself and fellow automaker Volkswagen focused on developing full level 4 autonomous driving technologies. Ford explained their justification in doing so when they released their Q3 earnings a few hours later, noting that not only were they shutting down Argo, but they were also essentially deprioritizing L4 technologies altogether, to instead focus on advanced driver assistance (ADAS) systems with internal resources.
Ford CEO Jim Farley justified this by saying on the company's earnings call Wednesday evening that "profitable, fully autonomous vehicles at scale are a long way off and we won’t necessarily have to create that technology for ourselves." The sentiments echoed those of a much younger and more tech-forward automaker CEO from just last week at our Disrupt conference in San Francisco.
While Rivian CEO RJ Scaringe did say the company was eventually aiming to introduce Level 4 autonomy, he also said that the plan is to focus first on L2 and L3 ADAS, with its existing shipping vehicles capped at L3 given their current hardware limitations. He did say that he believes L4 is actually currently possible for companies with the proper advanced hardware kit on cars -- with the caveat that those be geofenced to a specific location.
That brings us to the companies that are currently operating driverless vehicles on actual public roads, Waymo and GM's Cruise. Surely, if two (ostensibly) for-profit companies are already out there doing it, then it's going to happen, right?
The fact is that these existing services are extremely constrained in terms of geography and operating hours (though the latter is arguably a regulatory issue) and that seems unlikely to change at a pace that would make them ubiquitous in any reasonable time frame. Plus, the existing services face consistent, vocal criticism from residents who have to share the road with them.
At least Cruise and Waymo's vehicles are tuned for extreme caution -- possibly to a fault. Tesla on the other hand seems much more intent on hard-charging into a future where its so-called "Full Self-Driving" technology actually lives up to its name, with an expanding pool of beta users employing the tech on public streets and frequent software iterations that on at least one occasion have done more harm than good. Musk is also intent on stripping out as many sensors as possible from the Tesla autonomy hardware kit, probably in search of margins, under the misguided belief that compute, AI and optical input will improve and combine to act as a cure-all.
Musk strives to justify Tesla's approach on the regular, but it looks like he may have to do that explaining in more granular detail to the U.S. Department of Justice if they proceed with any action resulting from an ongoing criminal investigation the government branch is pursuing.
With even early pioneers striking a skeptical note, the time to consider the opportunity costs of shoveling money into the autonomy engine on autopilot is now. Argo AI was considered a leader with solid technological fundamentals by most experts in the field, so its shuttering is a strong signal not to be ignored. Meanwhile, I barely even scratched the surface on regulatory and public acceptance of any true ubiquitous self-driving, which will necessarily lag technological development — and likely by a lot, not a little.