IEEE: An interview with John Patrick
Saturday, February 5, 2000
An interview with
John Patrick is vice president for Internet technology, IBM Corp., and chairman of the Global Internet Project, an international group of senior executives committed to fostering continued growth of the Internet.
By Richard Comaford
Spectrum: What is the biggest challenge facing information technology (IT)?
We need information technology to be more natural.
I don’t mean just better traditional ease of use–competition among technology players will drive that. But to go from 100 million users to 100 billion worldwide, IT must become as natural as ordinary conversation.
Fortunately, there are emerging technologies that can do that. Language translation, combined with voice recognition, text-to-speech, and instant messaging applications, could result in transparent multilingual communication. Further, a person’s questions, regardless of what language they are expressed in, could be automatically routed to the most knowledgeable person within a specialized area. That person could answer in their native language and the answer would automatically be translated into the inquirer’s language. The whole process will appear to be very natural.
What are the challenges to implementing this technology?
We’ve worked on voice recognition for 20 years and, with a little training, the systems work very well; it’s not 100 percent, but it’s very close. Language translation could still be improved; it is not yet good enough for translation of a contract so it would stand up in court, but certainly it’s good enough for general interaction. The text-to-speech technology is quite good! You can add emotion to it, have a man or woman’s voice. So it’s not completely there, but it’s far enough along that you can see that it will be there.
As opposed to audio technology, what kinds of progress do you see with visual technology?
We’re making tremendous progress on the visual side. For instance, we have a technology called Cue Video that can take a one-hour program, analyze it, and represent it as a set of thumbnail images, like a storyboard. It can recognize when things are changing and can make a judgment about whether the change is significant enough to warrant a new thumbnail. Downloading the thumbnails requires very little bandwidth, and selecting thumbnails Nos. 4 to 8, say, plays those portions of video.
This permits searches of visual materials in two ways. One can search for a spoken word, find all occurrences of that word, and play the video segments where the word occurs, which is a very powerful capability. Then there is Qubic–for query by image content–whereby a video can be searched for, say, a brown plaid suit.
If you combine these visual technologies with natural-language technology, use of systems becomes much more natural.
‘THE INTERNET WILL CHANGE EVERYTHING–THE WAY TECHNOLOGY IS CREATED, DEPLOYED, AND THE WAY WE USE IT.’
If you had to choose one advance in information technology in the past century that will likely have the most impact on the way we live and work in the future, what would it be?
That’s easy, no contest; the answer is the Internet! It has extraordinary reach–the ability to connect everybody with everybody else. And it uses a standard approach; it’s the same everywhere. I can’t think of any other standard that is; electrical outlets are not the same, telephone jacks are not the same. Having been built on a globally adopted standard has allowed it to grow faster than any technology I’m aware of, including computers.
There have been many recent advances, but for the most part they have been because of the Internet. The Internet has enabled the collaboration that allowed those technologies to advance.
What is the biggest barrier to progress in information technology?
There are no showstoppers. But the area of policy is something we have to keep a very close eye on so that it doesn’t become a barrier. Collaboration is an extraordinary aspect of the Internet community; the question is whether policy development will also be collaborative.
One of the hats I wear is that of chairman of the Global Internet Project, a policy group consisting of 15 companies. We recently held a policy summit in Brussels, in collaboration with the Massachusetts Institute of Technology, and invited a number of leading experts from around the world to focus on potential issues three years down the road.
There was a strong consensus that there would be a collision of broadcast regulation and the Internet. The cost of high-quality production is plummeting (as evidenced by the popularity and success of low-budget efforts such as The Blair Witch Project). So anyone will be able to produce high-quality video, and when we get the bandwidth–which we are very clearly going to have through the competition between DSL [digital subscriber line], wireless, and cable–we will suddenly have digital video on millions of TVs. Look at the regulations that currently exist. In France, for example, 33 percent of the content broadcast must be in French. How do you do that on the Internet? The answer is, you can’t. But some people may try to apply existing laws to the new media. So the Project’s position is “Let’s start thinking of new forms of regulation…if regulation is needed.”
How will the Internet affect how we live and work in the future?
Basically, the Internet will change everything–the way technology is created, deployed, and the way we use it. I believe there is great potential here for achieving what I call “ease of life.”
Consider this scenario, for example. You receive an e-mail from the Department of Motor Vehicles that says, “Your driver’s license will be expiring at the end of this month. Please turn on the camera attached to your PC and click here if you would like to renew your license for the next three years using the same credit card you used the last time.” So you click and your picture is taken and sent to the DMV. Your biometric identification authenticates that it was really you, and a couple of days later you get an embossed plastic card in the mail.
There are many examples you can think of; it’s an end-to-end concept. There are a lot of neat things you can do on the Internet, and the Holy Grail is application integration. The Next Generation Internet applications will produce an ease-of-life standard. –R.C.
IEEE Spectrum January 2000 Volume 37 Number 1
© Copyright 2000, The Institute of Electrical and Electronics Engineers, Inc.