You can now make telephone calls, conduct videoconferences, listen to the radio and browse the multimedia World Wide Web over the Internet. It's all quite amazing, and it all consumes much more capacity than the traditional uses of the Internet, such as sending files or e-mail.
For instance, while most people send e-mail messages on the Internet, some prefer to "reach out and touch someone," using the new Internet Phone service, which allows them to make very cheap calls anywhere in the world. But a one-minute Internet call uses 500 times the capacity of a comparable paragraph of e-mail. And one minute of video covering that same paragraph uses 15,000 times as much capacity, the equivalent of sending more than 10,000 pages of text.
Can the Internet bear the strain of this high-intensity traffic? Information travels the Internet first-come, first served, so congestion from the new high-intensity uses already delays and interrupts other people. What will it be like in the future, as more people discover the exciting new uses?
Optimists believe we can keep ahead of congestion by simply adding more computers and lines to handle the traffic. While a popular solution to Internet congestion in the past, simply adding capacity won't work any more, for three reasons. First, with the government phasing out its Internet subsidies, we have fewer funds for expansion. Second, while the computers and lines that carry Internet traffic get cheaper to buy each year, they don't get cheap enough so that the Internet can grow faster that its traffic, which is doubling every year.
Finally, we have the new high-intensity uses waiting in the wings to soak up any extra capacity as soon as we add it. We also have a propensity to invent more such uses.
Extra money alone won't solve the problem of Internet congestion. But we do need to expand, and in the absence of government subsidies, that requires generating more revenue. Most Internet users (or more typically, their employers) currently pay a fee that doesn't depend on their use. They are free to create as much traffic as they like, without regard to the congestion it causes. We should raise revenue for expansion in a way that makes the heavy users aware of the congestion they cause.
At the same time, we should not discourage people from doing things that put little load on the Internet, such as sending e-mail.
The old first-come, first-served rule no longer works. Instead, we should prioritize Internet traffic according to the value users place on speedy delivery. Want extra speed for a message? You should make a voluntary payment to increase your priority. If you were willing to wait a second or two, you would pay nothing.
Frustrated Internet users could then do something about congestion. Now, for example, when an Internet Phone call gets garbled because of congestion, the only option is to wait until later to try again. But if callers could pay to increase their priority, then they could continue their conversation.
Audio and video don't work well if delayed, so those who want to use them are also more likely to want increased priority. But e-mailers, or others who don't mind a brief wait, can continue using the Internet without making priority payments. Those who put little burden on the system need pay nothing, whereas the high-intensity users who cause congestion would have an incentive to pay.
Allowing people to pay to overcome the effects of congestion would raise revenue for expansion, and at the same time make the Internet a reliable resource. For instance, a doctor who makes a priority payment could confidently use video to treat patients remotely. On the other hand, a college student unwilling to pay for the priority necessary to send a video to his sweetheart may settle for e-mail.
It would be a shame if we let traffic on the Internet slow to a crawl. Instead, we should take action to make sure the Internet realizes its potential as our information superhighway.