leonardo.gif (39048 bytes)


Leonardo Chiariglione is head of multimedia research at CSELT, the Italian telecommunications laboratory in Turin. He is father of the Moving Picture Experts Group (MPEG), which set the compression-decompression standards for video disks and digital television. He currently heads the Secure Digital Music Initiative (SDMI), which is developing a security standard that will protect the rights of everybody involved in the creation and distribution of music to everybody's mutual benefit. Despite some snags, Chiariglione is confident technical and political obstacles will be overcome.

Spectrum: In marketing Java, Sun projects an image of the electrical and electronics engineering communities as one big happy family, in which anybody can program in Java and play it anywhere. Yet one of the tools developed as part of the MPEG process--MP3--has turned out to have destabilizing consequences for one part of that community, the recording industry.

Ten years ago when we started MPEG, we were simply driven by the dream of, well, why don't we take these lousy and unwieldy analog signals and convert them into digital, and then they will become so flexible. You can do so many things. But then, in a separate world, the PC goes through different generations and all of a sudden Pentium comes and you can decode the MP3 files containing pirated music in real time. Meanwhile, 10 years ago we were using modems that transmitted at perhaps 2.4 kb/s. And then it becomes 10 or 20 times more than that, and, you know, these MP3 files can be transmitted over the Internet.

So, many fundamental developments in digital technology underly the MP3 saga. Considering those developments will continue apace, what big engineering challenges lie ahead?

The actual delivery technology will become opaque to the user: the user will simply say, give me 10 MB/s now. And then if it comes [in] broadcast [form], or if it comes interactive, if it comes whatever--you don't have to see that.

So transmission will become opaque. The same will be true of my computer platform: I don't want to know if it is Linux, if it is Windows, if it is whatever. The only thing that matters is: please give me access to computing power.

One technology that I think we are just seeing the beginnings of right now and will have the greatest impact in the next decade will be intelligent agents technology. The point is that today it's possible to inject simulated intelligence into computer codes. And, you know, because this code can be executed at virtually no cost nowadays, you can embed intelligence everywhere in the world, in the network, in your devices. This is something that is going to have probably more impact than digital computing, digital signal processing, and digital networking as such.

Bill Joy sees it as a big challenge for designers to develop appliances or devices that exploit all the potential of these fundamental developments

Well, this is one aspect, the extension of intelligence into things, intelligence in computer codes. You know, the way that you interact with the device is one example of how you can embed intelligence. But it goes beyond the simple interface. It's really that once this code knows about me, it can act on my behalf. And it can act on my behalf in dealing with other similar agents or codes representing other users.

You have pointed out that in the past, the consumer electronics industry has been very standards-based, and the information industry has tended to resist them, with the major players trying to take a proprietary approach. Do you think that now, in light of the Microsoft legal proceedings, that the kind of standardization process that you've been pioneering in MPEG is destined to now invade and take over the information industries?

Well, this is a very interesting question. And it allows me to give what will probably appear as a rather provocative answer. I believe that in spite of what people say, Bill Gates has been a benefactor of mankind.

Simply said, there were a number of processors, all doing more or less the same thing. He said, let's just pick one, and it's Intel's. Then, yesterday there were a lot of operating systems. Well, let's take one. All operating systems are the same. There is no practical difference. And then he went one step further. There were yesterday many ways of writing documents--the file format of a document, let's just take one.

Today, thanks to Microsoft, I and millions, hundreds of millions of people around the world, are at last able to exchange files which before they could not otherwise do because they found on their screen, "Unknown file format."

So what is my point? He, Gates, applied standardization principles by saying, one processing chip, please, one operating system, please. One type of Excel file, Word file, Power Point files and you name them.

Okay, so I think that Microsoft has proven for the IT [information technology] world the power of standardization. Do you think that the PC would be so pervasive today if we had had half a dozen competing operating systems, applications, as in fact we had, say, 15 years ago? The power of standardization has been proven--and if it has come about through somewhat unorthodox business practices, well, there are models of more open standards-setting procedures that give the same results with less hassle. --W.S.


IEEE Spectrum January 2000 Volume 37 Number 1

(c) Copyright 2000, The Institute of Electrical and Electronics Engineers, Inc.