Over on Slashdot, there's a vigorous debate about whether you can build a computer for less than $100, and what you'd get if you did.
While most of the comments definitely exceed my recommended daily allowance of geek, it's still a fascinating topic, if only as a data point for Moore's Law in action. As ally Charlie Stross reminds us, "[T]oday's million-buck supercomputer will be a desktop workstation in ten years time, a PC component in twelve years, and free with a cereal packet in twenty years."
I'm still not sure anyone has fully digested what that's gonna mean in terms of humanity's ability to solve problems (and create new ones).
What I find interesting about the idea of the $100 computer is that the thought was generated by a comment by Steve Ballmer, CEO of Microsoft.
A computer could likely be built for $100, but Windows XP certainly doesn't run on it. Over 10 years ago I started using the Internet on a computer that now I wouldn't pay $5 for and quite frankly I can't think of a whole lot I do with my computer now that I couldn't then with the exception of using it for things I used to use a TV and stereo for.
A computer which will run in a clean, dry environment with stable electical power can certainly be had for $100...
And a Palm Pilot is down around $70 new, and might be a much better fit...
Have a look at Nintendo's latest offering--the DS. It's being marketed as a video game device, but it's got 802.11b, two bays supporting memory of 80MB and 120MB respectively, two screens (one of which is touch-sensitive), a stylus, 16-channel audio out, and microphone support. All that for US$150.
Nintendo is opening the development on this pretty wide, and it's quite likely that someone will develop an email/feed reader/web browser/VoIP application for it rather quickly. This could be the device millions of lower-income people use to get on the internet, which is, in my opinion, the major reason to try to build cheap computers.