Hands-on lesson in getting a grip on tech
ANOTHER Tesla has crashed because the driver thought its self-driving technology could actually drive the car.
As we read stories about magical technology and then use the hyped-up products, we ought to keep in mind that the “magic” hits the market long before they live up to their promise, which in some cases they will never do.
The Tesla in Beijing, in autopilot mode, hit the side of an illegally parked car and kept going until driver Luo Zhen — who had taken his hands off the steering wheel — manually stopped it. It was a tough way for Luo to learn that, when he read and heard about self-driving cars, or even when he watched Tesla’s autopilot video (which tells drivers to grip the wheel at all times but shows the Model S changing lanes, taking curves and parking itself), he was essentially reading and watching sci-fi.
I’m not accusing Tesla of false advertising. The technology can do what the video shows it doing, but it can’t do it in every situation. That’s why the carmaker’s warning about holding on to the wheel is clearly articulated.
Nor was Microsoft really misleading customers about the ability of its Skype Translator to live-translate between Mandarin and English. It can do so when you speak slowly and clearly, avoiding complicated sentence structures, the way people do in the promotional videos.
But a Tesla cannot drive itself better than an experienced human driver can. Skype Translator can’t handle normal conversation the way even a middling simultaneous translator could. And Pokémon Go isn’t quite augmented reality.
It’s sad but true, whatever you think you’re hearing from starry-eyed tech writers, Silicon Valley marketers or even CEOs. When Apple boss Tim Cook recently said that “machine learning enables Siri to understand words as well as the intent behind them”, it was a forward-looking statement, not a promise to tomorrow’s iPhone buyers.
Two years ago, when muchhyped 3-D printing was proving a bit kludgier than neophytes expected, its inventor, Charles Hull, said: “Most of the stuff they talk about will happen someday — eventually. But there’s the here-and-now and the nearterm future, where a lot of that stuff is definitely hype.”
We don’t know whether, let alone when, autonomous driving technology will fully replace humans, or whether machine translation will work as well as the human kind.
We often pay to serve as testers for technology that is going exciting places for the engineers who develop it. And we expect instant gratification, although, intuitively, we know there’s no such thing in engineering. They don’t really deceive us: the warnings are there for those who will listen, and the makers and the hypers are rarely the same people.
It’s difficult for laymen to resist the hype. For those who need a reality check, though, there’s the “Hype Cycle”, developed by tech research firm Gartner. Technologies
The makers and the hypers are rarely the same people
aren’t brought to market when they can fully deliver on their promise but when they are at what Gartner calls the “peak of inflated expectations”. When it released its 2015 Hype Curve, autonomous vehicles, speech-to-speech translation and machine learning occupied the top spots.
Marketers figure the top of that cycle is the best time to offer tech to the masses. People who expect magic get disappointed, and the technology falls into the “trough of disillusionment”. But some of the technologies then make it up the “slope of enlightenment” to the “plateau of productivity”. Enterprise 3-D printing was halfway to the plateau from the trough a year ago, according to Gartner. Virtual reality was just climbing out of the trough.
This doesn’t mean we shouldn’t buy new tech when it’s being overhyped. But it does mean we shouldn’t expect much from it. — Bloomberg View