Twenty-five years ago the Internet was only 3 years old itself. At that time, less than 2,500 hosts were connected to the Internet and maybe 10,000 people used it regularly. Now there are more than 800 million hosts and 1.8 billion regular users.
But just looking at the numbers understates the impact the Internet has had on global communications.
I was an Internet user in 1986, and had been since 1983. I was a user of the ARPANET before then but I do not consider the pre-TCP/IP ARPANET to be the Internet since it used an inter-computer protocol rather than an internetworking protocol. The 1983 switch of the ARPANET from the Network Control Program (NCP) to TCP/IP enabled the Internet of today.
As an Internet user in 1986 I thought it was pretty cool, but it had a very heavy geek quotient. It never occurred to me that my mother would ever knowingly use the Internet. Too many seemingly magic incantations were required to get anything done. The magic required started to go away a few years later when the World Wide Web started being deployed in earnest. But the 'Net was still magic, maybe black magic, to a few important groups, in particular to telecommunications regulators, the existing telecommunications industry and the existing telecommunications standards development organisations.
The Internet protocol (IP) offered almost none of the features that the people in the telecommunications business in the mid-1980s felt were required for any useful communications protocol. IP did not offer any guarantees, quality of service, security or accounting. Governments were not involved in the standardisation of IP-related technology, and no one was in control. The obvious disconnect between IP, and thus the Internet, and anything useful for "good" telecommunications meant that most telecom players ignored the Internet until it was far too late for them.
In the early- to mid-1990s, I, along with Allison Mankin, managed the IETF effort that produced IPv6. During this process we asked many people for their for advice on the requirements for the next generation of IP. One person who responded was Hans-Werner Braun, then working for the National Science Foundation. He, quite seriously, said that we should assume we were developing the future of the world's telecommunications systems. Few, at the time, would have thought such a thing was possible, but Hans-Werner was right.
The Internet, and Internet-related technologies, have become the foundation of most modern telecommunications, both public and private.
It was not that long ago that editors of this column told me that their readers were not all that interested in so much Internet stuff, that was sometime in the mid-1990s. I took it to mean that the editors felt that something else, maybe ATM, would soon replace the Internet and, thus, would be the future.
Well, the Internet, to date, has vanquished all pretenders to the telecommunications throne. That does not stop some telecom standards bodies or carriers from working on technologies to "fix" all those things that are wrong with the Internet. Nor does it stop some would-be regulators from trying to guide the direction of Internet development so as to minimise the disruption of the technological or political status quo.
While the Internet has been amazingly successful, the 'Net is now brought to most people by a small set of what were once traditional telecommunications carriers, the very organisations that think the Internet needs fixing (usually in a way that provides the carriers with more control and money).
The future of the Internet, other than as seen by a few visionaries such as Hans-Werner Braun, has always been easier to experience than to predict. I see no reason for this unpredictability not to continue to be a feature of the Internet going forward.
Disclaimer: For most students, the results of their time at Harvard are also far easier to experience than predict. It appears, if measured by alumni donations, that their futures were more than fine. I trust that will also be the case for this thing we have been calling the Internet.