Is Google wasting its time with self-driving cars?

Google's self-driving car is a wonderful invention. But why is Google messing around with this? Shouldn't it be paying attention to problems with its core business?

Google this weekend disclosed its breakthrough work in self-driving cars, which have navigated 140,000 miles with very little human intervention (although trained humans are at the wheel to take over should the car decide on its own to do something dangerous, like, say, run over a pedestrian or kill everyone inside the cabin and eat their brains).

It's a fantastic breakthrough, but is automated cars something Google should be messing around with, wonders tech blogger Henry Blodget. The cars will take decades to deploy, and meanwhile Google has more immediate problems.

Why is Google spending the $10+ million of shareholder money per year the project consumes (15 engineers, plus drivers, plus the cars).

Isn't there something closer to its core business that Google could spend this money on?

Google Apps, for example. Google Apps are cool. But in many ways, they're still not ready for prime time. Wouldn't it maybe be better for shareholders if Google spent this money and focus on Apps instead of robot cars?

Or Chrome? Or Android? Or even search?

The myth of core competencies

TechCrunch's Michael Arrington takes Blodget to task in a post that basically calls Blodget a spoilsport.

But Arrington's post is unfair. It's true as far as it goes. Blodget is indeed skeptical that Google should spend its resources on automatic cars. But Blodget isn't arguing it's a waste of time. He's just saying that if automatic cars are a dream of Google co-founder Larry Page's, then Page should spin off a separate company to do it. That's what Amazon founder Jeff Bezos has done with is spaceflight dreams, he spun off as separate company for rockets, rather than having Amazon do it.

Fifty years ago, big companies had in-house blue-sky research departments -- AT&T Bell Labs is the most famous example. But the Silicon Valley business model beat those big companies, because it turned out that it was better to let startups manage innovation, while big companies kept doing what they were good at. Startups are more agile, and better able to compensate engineers for their work if an idea hit it big, Blodget says.

Blodget raises good points.

And yet this notion that companies have "core competencies," and anything outside those competencies is a distraction, is a bit of a myth. It's a way we explain failures after they happen. If a company tries a new venture, and fails, we look back and wave it off by saying, "That was outside its core competency."

Success outside core competencies

In fact, companies try new businesses all the time, sometimes failing, and sometimes succeeding. Take Blodget's own example -- It started out as an online book retailer, then branched out into other online retail. That appeared, at the time, to be outside Amazon's core competency; after all, you don't buy tube socks in a Barnes & Noble. And yet Amazon made it work.

Still later, Amazon went into business as a cloud computing provider. That, too, seemed to be outside Amazon's core competency -- Amazon is a retailer, not an IT vendor. And yet, Amazon seems to be making it work.

There are more examples: Microsoft started out selling software development tools, then branched out into operating systems, office suites, and Internet services. Apple started out selling desktop computers, then moved into mobile consumer electronics and media publishing. Telephone companies and cable TV companies went into the Internet service provider business after they'd been in business for decades.

The kernel of truth in the core competence myth is that companies are more likely to succeed where the new venture leverages old competencies. Amazon was already a retailer when it started selling underwear, and used its own global infrastructure as the starting point for its cloud services. Telephone companies and cable TV companies used the wires they were already managing to provide Internet services.

And in the case of self-driving cars, Google brings its massive computing infrastructure and data centers to the problem. The cars use detailed maps of the roads they're driving, augmented by realtime video cameras, radar sensors and a laser finder to see other traffic around them. Google has a head start on that information, and the techniques for gathering it, with its work on Google Maps and Google Street View databases. Google also benefits from the work of other researchers who've been addressing this problem for decades.

Even with that head start, Google might not succeed. The odds are against them. Self-driving cars are an enormous technology problem, that has resisted the attempts of brilliant minds to crack it. But if Google does succeed, the payoff is planetary, vastly bigger than even Google's current business. As Arrington notes: "In a hundred years who knows, Google may be thought of as a car company, not a search company. Crazier things have happened. It wasn’t all that long ago, for example, that Nokia was known as a manufacturer of rubber galoshes. If Blodget had his way, they’d still be at it."

Mitch Wagner

Follow me on Twitter
Visit my LinkedIn page
Friend me on Facebook

is a freelance technology journalist and social media strategist.

Copyright © 2010 IDG Communications, Inc.

It’s time to break the ChatGPT habit
Shop Tech Products at Amazon