Step into my laboratory

Step into my laboratory
By Margie Wylie
Staff Writer, CNET NEWS.COM

David Nagel will tell you that he's not a mad scientist but he's probably the closest thing to it.

Gray-bearded and genial, the six-foot something acoustics engineer would have a hard time getting cast as Dr. Frankenstein, but he is the scientist who will bring some of the intimidating technology of AT&T's now infamous "you will" advertising campaigns to market.

If you're one those people who've ever wondered aloud "what if I don't want to?" to those commercials, meeting Nagel, is reassuring. The new head of AT&T Labs comes to the job from Apple where he was most visible in his efforts to convince the FCC to offer free wireless Internet access to schools and universities.

Now Nagel is taking on another institution: AT&T. Besides filling some big shoes as the head of a laboratory started by Alexander Graham Bell, a laboratory that routinely turned out Nobel prize winners and inventions ranging from the telephone to talkies, Nagel is moving a better part of the labs to the West Coast where he'll be overseeing AT&T's move into computer-mediated communication businesses. His job is to make the telephone network friendly to new kinds of devices and applications, like the network computer or the flights-of-fancy kitchen and den computers imagined by companies like Diba.

Before Apple, Nagel worked for the National Aeronautics and Space Administration from its heyday in the early 1970's until shortly after its darkest days when the Challenger space shuttle exploded. At NASA, Nagel led the Human Factors team that worked on everything from spacesuit design and the first virtual reality goggles and gloves to such far-out projects as mind control for jets.

NEWS.COM: Why did AT&T choose you, a computer guy, to head up a telecom lab?
Nagel: AT&T is looking to be a much bigger player in the world of digital communications and applications and new services built on that. A lot of what's affecting the world of communications right now is coming out of the computing industry. The practices for creating new products and new services are being developed in the computer industry, and those practices are quite different from what they have traditionally been in the communications world. I suspect that in choosing me they were looking to tap into a bit of that culture: the strategies, the methods, the ways of working that characterize the industry in Silicon Valley more than in the industry of, say, the East Coast.

NEXT: Telecom Valley

Telecom Valley

What can the computer industry teach the telecommunications industry?
I'll give you an example. In the communications industry, the development of new services is traditionally done as a custom application. It's very analogous to the computer industry in the '70s: They use large pieces of equipment, each application is developed uniquely for each new service, and it's done laboriously. The tools aren't really good and so on. The great contribution of the personal computing revolution was the concept of a "platform," a standard hardware system and a standard software system that lots of people could build applications on top of, as long as the specifications were adhered to. In the communications business, you'd like to have platforms that not only AT&T, but other people, could build value on top of in order to provide services.

So with services like AT&T's WorldNet (an Internet service), telecommunications companies are moving into a new kinds of computer-mediated communications businesses.
I think the entire industry is, in some sense, headed that way.

For the most part, with the notable exception of PCs, the things that have been connected to the networks have been pretty stable for a hundred years. Until the last few years there weren't even that many PCs connected because there weren't that many people interested in the Internet. But the Internet has become almost a consumer phenomenon. There are probably 20 or 30 million more people connected worldwide to the Internet than just a few years ago. It is really becoming a general social phenomenon. I mean, literally, you can't sit in a restaurant without hearing the people at the table next to you talking about Internet. Sometimes it gets pretty boring in fact! But I think it's an inescapable fact that it has become part of the culture.

So people will be connecting all sorts of things to our network. We just introduced a new digital phone that can actually get messages off the Internet wirelessly. So that's an example of the kind of system that yes, we are very interested in. We won't be building them because we're not an equipment company, but we are a services company that will be supporting those. And part of my job is to determine what kinds of new systems we will support connecting to our network.

So tomorrow's telephone network will be more like, say, a Macintosh. We buy access to the network, but may run a variety of applications on that network?
Yes. I think that's going to be one of the things that will characterize the difference between today's phone system and tomorrow's digital communications systems: It will be much better thought of as a platform. And one of the tricks I think we have to work through is how we let other people, third parties basically, build innovation on our platform and deliver it to customers.

NEXT: Bigger isn't always better

Bigger isn't always better

I'm sure you probably have been watching with great interest the MCI and British Telecom merger. What does that mean to you?
It's rare that when you see mergers that things speed-up. So it takes a long time for companies of that scale to put themselves together in a way that is very effective in the marketplace. I think the short-term consequences could even be good [for AT&T] because it will slow them down in going after their traditional markets as they struggle with putting the companies together. So we'll see.

But the merger could be a boost for both of MCI and BT in the nontraditional markets, which really seem to be your focus.
So far, neither one of them have done so great in those new service areas. MCI started an Internet service provider service quite a while ago, and we have many more customers than they do in a much shorter period of time. So they have been notably unsuccessful in getting those new services. Will BT be able to help them? We'll see.

They seem like something of an odd couple. MCI won't put a penny into research and the BT Labs must have a phenomenal budget. Their scientists have been announcing a brain memory chip called the "Soul Catcher 2000."
There have been many brain chips and many ideas about using people's brains to control things directly. I guess it's a romantic idea. Many years ago when I worked for NASA, the Air Force was doing research on allowing pilots to steer aircraft by just thinking "left, left." It always struck me as a bizarre concept, because we have evolved sensory and response systems as human beings for millions of years and the idea that we could sort of engineer something in a couple of years that would do better than those frankly is pretty bizarre. So as long as they stick to that kind of research I'll feel even better. [laughs]

It will be interesting to see how that clash of cultures works itself out. MCI has always said, "we don't need labs, we'll just buy it. We can buy all [the technology] we need." They've always been very aggressive about that. It's just one more factor that might slow [Concert] down.

NEXT: You will...

You will...

I've noticed that labs more and more are being characterized and driven by profit above all things. Do you feel that sort of pressure to perform?
Look, it's an inescapable fact that everything is more competitive. I think whether you look at our lab or any lab, you're going to find a greater focus on being directly relevant to what the business objectives are. It can be getting technology to marketplace more quickly, putting greater attention on technologies or services or new products that yield a greater payoff earlier, or in some cases, it's providing a competitive advantage just in terms of knowledge. All those are valid roles for labs and I think ours is no exception. So it's not so much feeling pressure. Frankly, for me, it's more feeling the excitement of being able to make a direct contribution to business.

One of the things that I find in many of the people at least in AT&T Labs is that they are responding to that. They are feeling the same degree of excitement about being able to have a direct impact on many people's lives. I have someone that works for me that invented 800 service. About half the long distance phone calls that are made today are 800 calls. So here's one person who has literally affected hundreds of millions of people. That's really exciting for a scientist or engineer.

What about the role of serendipity though?
I think there's a myth that somehow serendipity only happens in a completely unconstrained, free environment. I think serendipity (as the name implies) is an unintended consequence or an unintended finding. That happens even in the most applied settings. The thing you have to do of course is be on alert for it, and take advantage of it, and exploit when you find it. But I frankly don't think there's any less likelihood that you're going to get serendipitous results in a more applied environment than in a completely unconstrained one.

This idea of research as a sort of an unconstrained activity--you give people money, you give them a site, you give them a facility, and they go off and sort of do whatever they think about and just follow their nose--is really a romantic and very inaccurate model of the way research is done in most places. The most basic researchers have a game plan. They can tell you exactly what they're trying to do and exactly where they're headed. And this idea that people run around in labs and white coats and sort of wait for inspiration to hit them is just nonsense. The most successful basic scientists have a game plan that stretches out for, in most cases, at least a few years and they know exactly what they're doing.

So you're really not just mad scientists running around?
Well we're certainly not mad scientists, but I think even the romanticized version of what research labs were ten years ago is not really fairly captured by that idea.

What about your own research Give us a little taste of the future, à la David Nagel.
One of the exciting things that I find about the immediate future is the potential for electronic commerce. I think for the first time we're beginning to see, particularly on the Internet, examples of people learning how to take advantage of this new medium, to do things differently. For example, one of my favorites (and it's actually a service that I use) is Amazon.com, the book service. Because it doesn't cost you anything other than a larger and larger database to have more and more books, you don't have to physically warehouse things, you don't have any scaling costs associated with being bigger and bigger and offering a larger and larger choice. So you end up with a service that's qualitatively, not just quantitatively, very different than any existing bookstore. I think that you'll begin to see other things like that electronically.

Of course we just introduced a whole series of new electronic commerce services that are useful both by customers. SecureBuy is a service that basically takes all the risk out of buying something over the Internet because we guarantee the purchase. It also makes available to merchants a new set of tools that they can use to hopefully create new services like Amazon.com.

Compared to the AT&T switched network, the Internet is like two tin cans and string. It's almost experimental at this point. How do you think it's going to change in the next five or ten years?
Well there's a lot written about that. I'm not so pessimistic. First of all the Internet has been around for a long time. We're not talking about something that was invented two years ago when the popular press started picking it up. The Internet has been functioning and doing quite fine and most impressively. Going from 10,000 to 10 million people is a huge increase in scale for any kind of a system. Frankly I think it's done a lot better than anyone would have guessed. And so I'm not so pessimistic that we're going to see some sort of an explosion of the Internet in which everything grinds to a halt and so on.

There are between 50 and 100,000 independent physical networks that interconnect today that form the Internet. And so we are all bringing on new capacity all the time.

When people start doing things that don't make a lot of sense, other people come in and change it. Yes, things may be a little chaotic in the transition periods, but you know, it all works through and stuff seems to continue to work.

So you don't think we really need to be vigilant as a society? You think that the market can take care of it?
Well I think we always need to be vigilant. I'm not a person for whom there's no role for government in any of this. In fact, I think there has to be. The Internet wouldn't exist if it hadn't been for the NSF or for ARPA, or for a variety of federal agencies who patiently supported it and helped create it and create the conditions and the market for it in some cases. So there's a strong and venerable history I believe, particularly in this area, of benefits from government involvement. The art is deciding when the government needs to step in, when they need to step out, and which things they should be most involved with.

I think the communications and information systems are among the most precious pieces of what we've been granted. Absolutely people need to be vigilant, but they also need to be informed. And they need not to be frightened. I think they need to be activists, be participants in all of this because if they are participants and they know what's going on, they have a much better chance of having a say in it and influencing it.

If I walked up and looked at you, what's the one thing I would suspect that you've never done?
Well that's quite a question! I think one of the strangest projects that I worked on, or worked within, was the idea of making an integrated circuit, an integrated system that was basically a Cray supercomputer on a chip. It was both ahead of its time and behind its time and it was not a great idea for a project. But that was probably one of the more bizarre ones that I worked on.