We live in the age of the digital packet. Documents, images, music, phone calls – all get chopped up, propelled through networks, and reassembled at the other end according to Internet protocol. So why not TV?

That’s the question cable giants like Comcast and Time Warner and Baby Bells like SBC and Verizon have been asking. The concept has profound implications for television and the Internet. TV over Internet protocol – IPTV – will transform couch-cruising into an on-demand experience. For the Internet, it will mean broadband at speeds 10, 100, or even 1,000 times faster than today’s DSL or cable. Online games would be startlingly realistic; the idea of channels would seem hopelessly archaic. Why not indeed?

So far, the answer has been inertia. But competition is a powerful stimulus. For years, DirecTV and EchoStar have been adding subscribers far faster than cable, so cable companies want something satellite can’t match. At the same time, voice over IP is enabling cable operators to poach phone customers from telcos. Combine VoIP, truly high-speed broadband, and totally on-demand TV – and you’ve got such a compelling proposition that the Bell companies figure the only way to survive is to do likewise.

IPTV is not to be confused with television over the Internet. On the public Net, packets get delayed or lost entirely – that’s why Web video is so jerky and lo-res. But private networks like Comcast’s are engineered, obviously, for reliable video delivery – which means IPTV will look at least as good as TV coming from digital cable or satellite.

It will be accompanied by another, equally critical change. Instead of broadcasting every channel continuously, service providers plan to transmit them only to subscribers who request them. In effect, every channel will be streamed on demand. This will free up huge amounts of bandwidth for hi-def TV and high-speed broadband. Add IP and you get interactive services like caller ID on your TV. And the system will be able to track viewing habits as effectively as Amazon tracks its customers, so ads will be targeted with scary precision. Put it all together and you’ve got television that’s as intensely personalized as 20th-century broadcasting was generic.

But that scenario is a good five years out. First there’s a lot of upgrading to do. The Bells have the worst of it: Their copper lines max out on current-generation DSL, never mind TV. SBC is testing an advanced form of DSL that promises 7 times the bandwidth it delivers now. Verizon plans to spend billions to provide its 30 million customers with direct fiber connections offering nearly limitless bandwidth. Cable companies have already upgraded their networks, at a cost of some $85 billion, but it will be 2007 before they complete the transition from analog to digital and some time after that before all their customers get IP-addressable set-top boxes.

More here.