This is my proposal at the meeting where we planned NREN in February, '87. It's turned out to be an accurate projection. ARPAnet was overloaded as the country's research network. The proposal was to start from 56 kilobits and to get what we had working. To get to T1 quickly by the late '80s. And then to rapidly go from T1 to DS3 or 45 megabits in the '93 time frame.
What you can see from this, is that a factor of a thousand is really what made the difference. What you get with a very fat pipe is not the ability to send bogus data and pictures or video. But you get the ability for a lot of people to send a lot of small messages and get them instantaneously. This is what makes Internet 2.0 work in 1995.
We still have another factor of a hundred or so to go to adhere to the projection. We need probably another factor of one thousand to do the kind of things for Internet 3. The optical gigabit links called for research, all these others were more straight-forward. The Internet task force worked very hard. And Internet is the most impressive engineering I know, but done as what might be called chaotic engineering. I think its the greatest network engineering that's come down the pike.