It's long been an article of faith for many mobile enthusiasts that mobile phones are going to become the dominant means by which the human race deals with the internet. Lately that idea has been echoed by more mainstream voices:
"There are over one billion people in emerging markets who will never access the Internet using a PC."
-- Adobe CEO Shantanu Narayen, on stage at CTIA, September 2008
"Most new internet users will be in developing countries and will use mobile phones."
--The Economist, September 2008 (link)
"Mobile phones are the future of computing, and they are ideally suited for accessing Web services....(The mobile phone will be:)
-the device most likely to subsume the PC's computing and informational dominance
-the most functional and accessible device for conducting Web search
-the natural gateway to Web 2.0 platform applications and services"
--E-Week, October 2008 (link)
I have a lot of respect for those sources, but in their enthusiasm for mobile technology, I think they have made two big mistakes:
--They've assumed the internet is a thing, and
--They have forgotten about Moore's law.
The internet is not a thing
It's a collection of standards for data transport, storage, presentation and so on. People do an incredible range of tasks that take advantage of the internet, some of them well suited to a mobile phone and some of them not. Creating documents and graphics that you want to share online, browsing content, and writing online comments all are moderately to enormously easier when you're using a PC with a keyboard, mouse, and larger screen. They also benefit from having large amounts of local storage, which you don't have on a mobile phone.
The idea that people in the developing world won't want or need the benefits of a larger screen and keyboard is patronizing. It assumes that they'll be content to be second class citizens for many Internet services permanently.
If they can't afford a PC, of course they'll do what they can with a mobile phone. But that brings us to the second assumption -- who says they'll never be able to afford a PC?
Moore's Law lives
As I've written in the past, some people in Silicon Valley worry that we're starting to run up against physical limits on the growth of computing performance (link). Although that may or may not turn out to be an issue at the high end of the market, at the low end no such barrier exists.
I'm told by friends on the manufacturing side of things that the production cost of a fully-equipped ultra-mobile PC or netbook (what we used to call a mini-notebook) is now around $200. The street price for basic models is $282-299, a drop of about 17% in the last four months (see here, and here). Many of the key components in UMPCs, such as the screens and optical drives, are also used in DVD players, which means they're being manufactured in large volumes, driving rapid price erosion.
UMPCs aren't perfect -- the keyboards are very cramped, and the screens display text so small that they can be uncomfortable to read. But they are far, far better than a smartphone for many computing and internet tasks.
I am not saying that PCs will become affordable for the world's poorest people anytime soon. But let Moore's Law continue to chew on the UMPC, and I think a PC will soon be within the reach of a working-class family in much of the developing world.
The most likely outcome is that most people who can afford a mobile phone and service plan will also be able to afford a small PC if they want one. My guess is that they'll use both devices in the same way you and I do -- the phone will be better for some tasks, the PC for others. The idea of using one to take over for the other will seem silly, kind of like using a hair dryer to cook dinner.
Someday in the distant future, of course, we'll have smartphones with flexible screens and fold-out keyboards that can fulfill all of the functions of a PC. At that point, the line between your PC and your phone will blur, and you'll be able to say that your phone has taken over your PC (you'll also be able to say that your PC has taken over your phone). But after watching the lethargic pace of change in mobiles over the last ten years, I think we've still got a long wait for the merger of phone and PC. Besides, integrating more features generally raises prices, so the merger of phone and PC will happen first at the high end of the smartphone market. PCs will be dirt cheap long before they merge physically with phones.