Cuban has a good point. The flip side...
...of course is that in a world of limited bandwidth, "fast lanes" can only exist alongside "slow lanes." This will become more apparent as we move forward without doing something about aging internet backbone infrastructure. What happens when the major content providers all pay for "fast lane" status, but the cost is prohibitive for small start-ups to do the same? In that case, fast lanes can become a choke on innovation, rather than a boon to it.
I don't think anyone will deny that the internet's move to a neutral model in the mid-90s was good. That doesn't mean that an entirely neutral net is the right model moving forward, and it's certainly possible that the internet will always work best - neutral or not - when it is relatively unregulated.
I think one of the main problems with this debate is that we often speak about two different issues as though they are one in the same. Many large telecoms own both part of the internet backbone and direct-to-consumer ISPs. Data transit on the internet backbone is already subject to myriad agreements between companies regarding transfer allowances, cost sharing, peering, etc. This is a good thing, so long as no one entity (or small number of them) own a majority of the backbone infrastructure, similar to how the rail system used to work. Although some proponents of net neutrality would want an entirely blind network, this isn't really where the debate exists for most consumers and content providers.
The real debate has to do with last-mile neutrality, with ISPs limiting bandwidth usage at the consumer's point-of-connection based on protocol, source, content type, whatever. This worries people because it reminds them of the days when AOL, Prodigy and CompuServe controlled what content could even reach consumers. I was watching old NCAA highlights last night, and during the 2002 Tournament, CBS was promoting "AOL Keyword: CBSSports," or whatever it was. Some 7 or 8 years after AOL opened up sockets to 3rd party browsers, and the very definition of the World Wide Web was still shaped by that company's dominance and restricted-channel model of content provision. This is the real battle, because there is no legitimate reason for ISPs to choke bandwidth at the end-point. Users already pay a monthly fee based on a maximum bandwidth allowance; they should be able to use that bandwidth for whatever content they see fit (provided it's legal, of course). If ISPs can't afford for consumers to actually use the bandwidth they are already paying for, then there is a problem with the pricing structure itself.
goat