Intelligent Machines

Waiting for Google Cube (or a PC Like It)

The rumored “Google PC” announcement has not materialized, but the time for a networked computer with an alternative operating system is right. Here’s why.

Earlier this week, the buzz at the Consumer Electronics Show in Las Vegas was that Larry Page, Google’s cofounder and president of products, would be unveiling a low-cost, Google-branded, networked PC. The device even had a nickname: the “Google Cube.”

The rumor, first circulated by pundit Robert X. Cringely, and expounded upon by the L.A. Times this week, centered around a computer running a nonMicrosoft operating system and featuring only a minimum of native, or desktop-based applications, relying instead on applications available on the Internet. Most of the intelligence and applications for the unit (word processing, spreadsheets, Internet, etc.) would reside online.

Google quashed those rumors yesterday, though, and reports in the Wall Street Journal and elsewhere now indicate that the company will instead announce a downloadable video service and “Google Pack” – a handful of media-centric applications.

But, despite its recent no-show, many people believe a device like “Google Cube” will arrive someday, from Google or another company. And despite conventional wisdom that launching a low-cost, commodity-like PC is a fool’s errand, there’s a lot that makes sense about a cheap, network-centric computer, whether it’s a Google product or not.

Certainly, the strategy of selling low-cost – or no-cost – PCs to jumpstart interest in more lucrative programs is not new. Nor has it been particularly successful. Larry Ellison’s mid-1990s Net PC was one of the first mainstream efforts to put the intelligence in the network and return to an era of “dumb terminals.” Later, in the late 1990s, Internet service providers such as Empire.net gave away PCs in exchange for lengthy memberships. In 1999, Free-PC, a company launched by dot-com incubator IdeaLab, gave away PCs in exchange for user’s agreeing to have the margins of the computer screen occupied by banner ads. None of these attempts survived the dot-com implosion.

When asked in 2002 by Wired News to name his biggest mistake of the bubble-era, Bill Gross, the erstwhile CEO of IdeaLab, responded: “FreePC was one of my worst ideas.” But showing prescience even in the depth of the post-bubble fallout, Gross went on to envision “some point where there is a device that will give you some access but will be advertising-supported.”

That time might well be now. Recent advances in software, combined with a resurgence in online advertising and widespread adoption of broadband access, make it so. The most lucrative advertising online today, on a pure cost-per-thousand basis (the basic unit of determining how much a single online ad costs), is the ads surrounding video snippets. This form factor wasn’t viable in the late 1990s – broadband connections didn’t reach enough homes. And until around three years ago, the online ad market was still struggling to reach its pre-bubble highs.

The most important factor, though, is the recent and rapid improvement in Web software. Thanks to these innovative applications, a non-Microsoft, non-Mac computer could operate in either one of two ways. It could be a true “network device,” relying on Web-based applications for all its needs and shipping without a full-featured operating system. Today, browser alternatives such as Firefox and Web-based applications such as Writeboard (word processing) and Basecamp (project management) are proving that an application can be as useful when based online as on a desktop. Across the Web, a new wave of software development is rising up, using technologies such as AJAX (Asynchronous Javascript and XML) that bring new functionality and desktop-like performance to Internet applications.

Google’s Gmail is an excellent example of this trend: it’s a Web-based email program; but unlike Hotmail, which must refresh a page whenever you close a message or want to reply, Gmail’s foundation in AJAX allows it to appear as if the program exists on your desktop.

The second option for such a PC would be to ship it with an alternative operating system, such as a desktop version of Linux, and for it to run non-Microsoft applications, such as Sun Microsystem’s Star Office – a counterpart to Microsoft’s Office program. This, too, is a general area that has grown in the last couple of years.

This scenario does not deny that Microsoft is still a monopoly in office applications, and its browser, Internet Explorer, is the hands-down leader. The biggest challenge for alternative products in the marketplace is “consumer demand,” according to Kevin Carmony, CEO of Linspire, a desktop Linux company. “It’s very difficult to educate consumers after 20 years of Microsoft that there’s an alternative. They’re brainwashed into thinking there’s one choice.”

Enter Google. If there’s one company with the deep pockets and high regard to make a legitimate run at changing consumers’ computing behavior, it’s Google. The company is an odd bird that appeals to engineers, investors, and grandmothers alike. If history is any indication, however, this alignment of factors won’t last that long. While the announcement of “Google Cube” isn’t coming today, Google and its ilk should not let the moment slip away to reinvent the PC.