Google isn't shy about doing novel, controversial, and occasionally even quixotic things, and there's a chance that its latest venture, Chrome OS, will turn out to be all of these things. Chrome OS may turn out to be a failure, but if that happens it's probably only because it's ahead of its time, and Chrome OS may well change computing forever. Not because of what it is but because of what it represents.The computer you're probably reading this on (and the one I wrote it on) are only modernized versions of a system that was first sold when the worldwide web didn't exist and when the Internet was just a small research network. Suppose that, when the first PC was being conceived, we'd had broadband networks with network storage, computing, and applications available? What kind of system would we have had? There's a name for it--"terminal." The broadband Internet translated to the '70s might have solidified the concept of mainframe interactive computing. Scary, huh?
Universal broadband in the '70s isn't an option today, but that doesn't mean that it won't have a profound impact. The paradigm of the mainframe has been replaced by the paradigm of personal computing (what's "PC" stand for anyway?) and that's what now is colliding with the ultimate global network. There was a time when bits were scarce and expensive; now they're increasingly cheap and quickly becoming universal. Any radical change in the price of a commodity changes the market, and this change is going to transform computing completely.
Freed of constraints on the cost and performance of network connectivity, things beyond very basic and truly personal computing are going to get sucked into the cloud. For enterprise users this will be visible first as a polarizing of features in personal productivity applications like word processors; you'll have "temporary" document storage and abbreviated features you can exercise on your netbooks or whatever, but the real horsepower will be in the cloud.
We'll also be seeing a new model for virtual desktops, one that has two distinct layers. There will be a GUI designed for every type of client device, optimized for what that device can show and what kind of realistic data interactivity it can support. There will then be a mapping of every personal application to each of these GUIs. When you have a device that's got a good keyboard and local capabilities, it will unlock one set of application features, while a more limited device will support only a limited subset. The virtual desktop future will embrace smartphones (Android, perhaps?) or big flat panel displays; little processors or multi-core behemoths. As long as you're prepared to accept the functional limits of devices, you can use pretty much anything to access your applications.
This kind of world is hard to visualize when you've grown up with the notion of a personal computer. If you look at the early commentary on Chrome OS you see comments like "no local storage" or "limited offline utility," and all that proves is that we can't easily shed our old image of personal computing. The whole idea is to reduce local storage, to reduce local applications. Moore's Law has made computers cheaper but no less complex. The limiting factor in the spread of sophisticated applications is the need to support them and their user base. Centralization lowers the bar for distribution of high levels of functionality.
Group editing of documents today is complicated by the fact that there's a copy of the document for everyone in the group. In a world where the documents are hosted, collaborative editing doesn't require synchronization of multiple "spatial" versions on multiple systems, only of multiple chronological versions. Does that sound like Google Wave? No coincidence.
All this will impact business voice communications, I think. If you leave the social aspects aside for the moment, business conversations are about creating a collective decision, normally based on some set of common information and represented by some final document. As we increase our ability to link the inputs (the sources) to the outputs (the presentations or reports) in some serial, hosted, process we can expect to see people working more in their own way in their own time rather than trying to zero in on a common time for a conference call or even telepresence meeting. And how much social chatter today is done with messaging/email versus voice, anyway? The new generation of workers are just as comfortable with text as with voice-maybe more so.
A future collaborative world, a world of unified communications, may end up being created by something other than pulling voice onto desktop systems through APIs or providing voice readout of emails. The business world, and even our personal tech-supported worlds, are giant ecosystems pulled in different directions by major changes in technology, and nothing can compare to the broadband revolution as a major change. Cheap bandwidth literally changes everything because it erases the barrier to centralization of complexity for easier management. Thus, that centralization is inevitable.
Whether Google is trying to drive this, sees it as a kind of primal market force, or simply sees it as a convenient path to greater ad revenue is impossible to say. They are aware of all of this. Because Google was born in the online world-unlike IBM or Microsoft--it's not committed to defend a legacy of personal computer evolution. Since all it takes is one attacker to create a war, Google's awareness of the future is already forcing its competitors to take another look at their own strategies, and that will only accelerate the network-centralization of our world.
Sometimes you can get so tied up in evolution that you miss the revolution completely, but not for long. It looks to me like 2010 is going to be a very interesting year.
