Mark Cuban writes: "Craig Moffet of Bernstein Research was asked to testify before the Senate Commerce Committee on the subject of Net Neutrality. His comments were of course right on the money. The interesting conclusions that can be drawn from his testimony are just as relevant to the discussion of the future of media on the net as they are to net neutrality. ¶ Craig sites facts and figures that should make anyone who believes that the net as alternative to TV is just around the corner, or will happen this decade for that matter, rethink their position. ...¶... Craig is right. The last mile into our homes wont have enough bandwidth to support all that we will want to do via our internet connections at home. There is no Moore's law for bandwidth to the home. There is a huge misconception that bandwidth will just continue to experience unlimited expansion for every broadband household. Its what we are used to with hard drives, processors, all technology. It gets faster, cheaper, bigger. Thats not the case for the next decade with bandwidth. ¶ The net result is that TV is going to be TV, delivered like TV for a long time to come. (I consider IPTV to be regular TV). There won't be enough bandwidth for it to be any other way. ..." Link: Blog Maverick.
My standard for posting items to this blog is that they present thoughtful opinions, not necessarily whether I agree with them in whole or in part. When I disagree with one, I seldom push back, but I'm going to do that here. I think Cuban and Moffet are missing some elements that make use of much more modest Internet pipes useful to a large number of consumers.
- They don't really distinguish between real time and non-real time use, and their argument holds only if all or most programming is required to be delivered in real time. Many consumers (users of TiVo and other DVR/PVR devices, Netflix subscribers, cable VOD subscribers) are already proving that non-real time delivery is valuable. If you think of Netflix as a sort of broadband pipe, it's delivering a 1-2 GB movie in 24-72 hours, depending on where you live and when you order it. An HD movie would be delivered in the same time.
- They don't consider the impact of storage, which is a linear trade-off with transmission capacity -- and probably much cheaper to capitalize. Today's 100-GB TiVo should be 1.6 TB in five ticks of Moore's law (ca. 2012) and 51 TB in ten ticks (ca. 2020). Compression technology will continue to make that space more efficient. Although too young to be proven successful, Moviebeam has the right idea -- a big hard drive, preloaded with movies, that is "trickle charged" with a few new movies over an extremely modest connection (sub-dialup spead over analog television datacasting).
- Finally, they don't consider the impact of advanced (P2P) networking technology. In P2P systems, the more nodes that store a given file, the faster the download and therefore the cheaper the network that can support it. Popular television programming and movies are at the core of their argument about capacity problems, yet these files, which are at the fat end of the Long Tail, will by definition reside a lot of places and should be ideal for the benefits of P2P networking.
Added 3/20/2006: Be sure to read Chris Duncan's thoughtful comments (also posted to Cuban's weblog) below as well as those of some others (including Cuban's reply to Duncan) that have been added to Cuban's original posting. --Dennis