Self-driving cars are a stupidly-expensive distraction

As you know, QuadRanting is no fan of autonomous vehicles. It’s not just that they are infinitely more difficult to achieve than their over-excited fans would have you believe, it’s that no one has really come up with a convincing reason why we need them.

Of course, since self-driving cars (SDCs) are the greatest of techno-narcissism’s current grands projets, the requirement for something as dreary as a worthwhile raison d’etre for them is considered to be an irrelevance. They’re exciting and theoretically achievable so let’s pour billions of dollars into them!

In the same way that the word ‘blockchain’ magically enables unscrupulous outfits to relieve gullible fools of their money, any story involving ‘self-driving cars’ is fully guaranteed to gain wholly uncritical media coverage, even in outlets that should know better. So when we learned that fully-autonomous vehicles will be on the UK’s roads by 2021, it was no surprise to hear the news not from the lips of Ricky the Magic Pixie but from those of the Rt. Hon. Philip Hammond MP, Chancellor of the Exchequer.

You have to ask why the automotive industry – huge, highly regulated and quite responsible at the operational level, though not existentially – would want to play silly buggers in this way. The answer is that by and large it doesn’t.

There is a smidgen of truth in the idea that the big automakers fear they’ll be put out of business by Google and other tech giants flexing their artificial intelligence muscles. But the manufacturers know they have more pressing threats to their future to worry about than toy cars adorned with $100,000 dollars-worth of sensors that need to be connected 100% of the time to a multi-billion dollar IT infrastructure that itself is guaranteed to be ‘up’ 100% of the time. Yup, 99.99% won’t be good enough. Especially when even Chris Urmson, the former Chief Technical Officer of self-driving cars at Google, recently said it’ll be at least 30 years before anyone achieves a truly self-driving car. So 2050, Mr Hammond, not 2021.

Above all, motor manufacturers understand that even relatively minor moves (in comparison to putting a fully autonomous vehicle on the roads), like developing a new hybrid drivetrain, are immensely complex global undertakings that span international supply chains, multiple vehicle regulators, varied climates, traffic rules and highway design protocols sales channels, and so on. Agreeing something as seemingly insignificant as a global standard for an electronic interface that allows different components to talk to each other can take many years.

It’s obvious that a lot of people are mentally constructing castles on the sand in order to clutch at pies in the sky.

But something has caused Wired.com, hitherto one of the most shameless boosters of all things autonomous, to inject a note of realism into its discourse. Perhaps someone there had a butchers at the rising tide of SDC stories and thought, “Hmm. What does that remind me of?”

Hype-cycle

In Before Self-Driving Cars Become Real, They Face These Challenges, Wired quotes Bryan Salesky, who heads up Ford-backed autonomous vehicle company Argo AI, saying (emphasis added):

“Those who think fully self-driving vehicles will be ubiquitous on city streets months from now or even in a few years are not well connected to the state of the art or committed to the safe deployment of the technology.” Medium.com

Take note of the comment on safe deployment. Achieving Level 5 (total) automation will be incredibly difficult. From the Wired piece:

”Technology developers are coming to appreciate that the last 1 percent is harder than the first 99 percent,” says Karl Iagnemma, CEO of Nutonomy, a Boston-based self-driving car company acquired by automotive supplier Delphi this fall. “Compared to last 1 percent, the first 99 percent is a walk in the park.”

Do the math. He’s saying that when the SDC-ers get 99 percent of the way to Level 5, they’ll be less than half-way to actually putting the technology safely on the road. Unless the rate of development goes exponential, it will take decades to iron out the final bugs between Level 4 and Level 5. And while we’re at it, let’s mix-in the warning from Salesky, above: developers – or countries – that try to rush the process will kill people.

Wired calls Iagnemma a killjoy. Its piece concludes with the obligatory happy ending:

The good news is that there seems to be enough momentum to carry this new industry out of the trough and onto what Gartner calls the plateau of productivity. Not everyone who started the journey will make the climb. But those who do, battered and a bit bloody, may just find the cash up there is green, the robots good, and the view stupendous.

I suspect the phrase ‘battered and a bit bloody’ will sooner rather than later prove rather too literal as far as some passengers and bystanders are concerned. Expect it to come back to haunt Wired.

‘Stupendous’ looks misplaced as well. What the plateau delivers is well short of what peak hype promises. A Gartner curve adapted to SDCs might look like this:

Autonomous-Vehicle-Hype-cycle

I’m not sure we’ll exactly see a trough of disillusionment. People, especially in the media, are deeply wedded to the belief that everything is getting better and that technology can deliver all our dreams at the flick of a venture capitalist’s quiff. Unless or until there’s a global depression deep enough to scour right through the layers of hype, sheer techno-optimism will ensure the SDC-ers have sufficient Other People’s Money to burn.

Nevertheless, the climb up from reality setting in to the plateau of productivity will be a lot, lot longer than implied by the Gartner model. And the economic headwinds will be getting stronger all the way up. The market for any car, let alone a super whizz-bang expensive SDC, is shrinking along with the dwindling buying power of ordinary people.

Sounding somewhat desperate, one SDC start-up is aiming at retirement communities – clearly hoping that the pension ponzi doesn’t pop before its founders cash out. Another aims at disabled users. That’s a better idea, in that there’s actually a problem there that an SDC could be the solution to. But again, it’s a niche market that doesn’t justify the stratospheric cost of attaining Level 5. There are undoubtedly better and cheaper community-based solutions for disabled travellers.

All in all, SDCs are industrial civilisation’s equivalent of brilliant plumage in the animal kingdom or exotic blooms among plants. But unlike natural showing-off, which is developed to be sustainable over millennial timescales, SDCs are no more than hubristic adornments at the apex of a system that’s already running out of cheap energy and starting to drown in its own effluent.

When the history of the 21st century comes to be written, SDCs will undoubtedly feature on the list of Stupidly Expensive Diversions from Attending to Real Problems.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s