We did it; we finally broke the Internet. You shouldn't see Time Warner's capping home users' bandwidth and AT&T's momentary, panicked foray into banning Slingboxes solely as examples of ISPs greed. See them also as signs of ISPs' deep abiding fear.

As we've become a high-def, two-way, video-on-demand society, we're overloading the wires and, especially, the airwaves along which our data travels. The result is going to be a three-year slowdown of the Internet experience, with more expensive connections and innovation turning toward efficiency rather than extravagance.

I'd love to pile on the bandwagon claiming that new Internet restrictions and fees are due simply to greed. And when we're talking about home connections, greed does seem to play a big role; it's no coincidence that the most-used high-bandwidth Internet application, video, competes with ISPs' cable TV and proprietary video products. There's a new technology, DOCSIS 3.0, that cable companies are starting to use to speed up home connections, but cable firms are using it also as an excuse to increase their rates.

A recent New York Times story suggested that local cable providers are dragging their feet on equipment upgrades so they can use congestion as an excuse to stop Internet video in its tracks. They claim there just isn't demand for more capacity, which is, of course, utter nonsense. If there wasn't demand, they'd be lowering rates, not raising them. Our dysfunctional franchise-monopoly system puts us at the mercy of the cable providers, for now. And barring serious competition, the cable providers would rather charge by the megabyte than upgrade their networks.

3G wireless doesn't provide much of an alternative. I've seen the terrified looks in the eyes of wireless company executives. They're genuinely worried that they can't carry much more traffic. 3G networks are much more fragile than you think. AT&T's network couldn't even handle a bunch of SXSW convention attendees with iPhones. Now imagine if they were all trying to live-stream video using Qik. I've had wireless executives tell me that they set data plan prices high specifically to keep people off their networks—because if they had too many more heavy data users, the networks wouldn't be able to function.

Monopoly economics on the wire-line side and inadequate networks on the wireless side means we're not going to see cheap, unlimited Internet over the next few years. The result of this mismatch between applications and infrastructure is going to hurt everyone at the start, but don't underestimate clever technologists. Bandwidth will become a currency. Web pages will get more basic; marketing bloat will decrease.

Twitter is a big winner here. Text is a big winner. Text is very efficient. Wordpress, Blogger, Crisp Wireless, Usablenet Mobile, Opera Turbo—and anyone who makes the Web smaller or more efficient—are winners in the new math.

People who develop new, efficient media codecs (such as On2 Networks and PacketVideo) could be big winners. Adobe could be a loser, because the folks who stream video will be big, big losers. YouTube? Loser. Hulu? Huge loser. And, of course, that will make the cable companies quite happy.

There's a light at the end of the tunnel, if a somewhat dim light. Four wireless carriers are all building fourth-generation LTE wireless networks, which will take some of the strain off the current 3G systems. We won't see LTE in wide deployment until 2011, and it too has its limits. But at least the limits are better than 3G's limits.

For fixed installations, we're in the middle of a painful, expensive process of installing new infrastructure to compete with the cable companies. Verizon's FIOS costs hundreds of dollars per home to install, as The Times noted. Clearwire is dropping millions of dollars to set up a nationwide WiMAX system running at about 6 megabits per second—a process so expensive that many people are unsure whether Clearwire can afford to finish the build-out. And once the cable companies have real competition, they may finally price their new DOCSIS 3.0 systems realistically rather than extortionately.

It's hard to figure out the public policy that would encourage competition here. Getting the FCC to crack down on cable companies would lower prices but would just create a slow-moving, sclerotic, state-controlled monopoly to replace the sclerotic privately owned monopoly. Regulation is great for prices, but it's unlikely to encourage innovation. We could throw public money at broadband build-outs, but last I checked we didn't have much public money to throw anymore.

No, I think we're entering an era of efficiency—of doing more with less bandwidth. That doesn't mean we're going to stop innovating. It just means we're going to do it with fewer bits.