Keeper of the Domain.

Herb Schorr's Information Sciences Institute is the quiet force behind many key innovations for the Internet

TODAY, we take things like domain names and other elements of Internet architecture for granted. But where did they come from? Bill Gates? Steve Jobs? Nope.

In fact, much of the technology underlying the Internet has sprung from a little-known, behind-the-scenes group based in Marina del Rey called the Information Sciences Institute of the University of Southern California.

Under the direction of Executive Director Herb Schorr, researchers at ISI are working on the technologies of tomorrow, most of which could easily be translated to a marketable product for businesses or individual consumers.

With more than 325 researchers, students and staff on both coasts, 151 was formed in 1972 to manage and research the ARPAnet, the precursor to the Internet. Every time you type in a domain name like yahoo.com or MSNBC.com to go to a Web site, you're using a system developed by ISI.

Most of the group's current work has to do with technical aspects of Internet architecture, such as file system infrastructure or new ways of doing teleconferencing, and it maintains one of the nation's largest artificial intelligence research centers. The organization is supported by contracts and grants from the federal government and from a variety of private corporations including Intel, Microsoft, Hewlett-Packard, IBM and Cisco Systems.

Prior to joining ISI, Schorr spent more than 30 years in various engineering and management positions at IBM. His specialty became working in groups that developed new technologies. He's run ISI for almost 12 years.

Schorr met with the Business Journal in his corner office with a panoramic view of Marina del Rey on a recent afternoon, to discuss the organization that does so much in relative obscurity.

Question: For those who've never heard of it, describe ISI.

Answer: We build systems, in contrast to most universities. We're sort of between a corporate laboratory and a university in the sense that we do good research by building something new.

Q: What's the origin of ISI?

A: We're a spinout from Rand. The ARPAnet was starting, and ARPA (the Advanced Research Projects Agency) wanted a place to help run the Internet and do research on it that was less expensive than the normal defense contractors. Universities didn't have enough continuity, with the grad students who come and go. Rand had to make a choice between computer science and the studies you normally think about - the place was ran by economists. So a small group spun out. They decided they wanted a university affiliation. To make a long story short, both universities here said yes, but UCLA said it would take them a year and USC came back in a few days and said, "When can you start?"

Q: What was ISI's role in the ARPAnet, the precursor to today's Internet?

A: We ran the ARPAnet for a good time. In the beginning there were big timesharing computers in the 12th floor of this building. We developed a lot of the protocols, particularly the domain-name system. That was invented here. You know what ICANN is? That's in this building as we speak. It essentially came out of the domain name system. We ran IANA, which was the predecessor to ICANN. We spun LANA out as it became more international. It seemed it shouldn't just belong to us; it needed to belong to the world.

Q: Did ISI pioneer anything else we see on a day-to-day basis?

A: There's a system here called MOSIS. It allows people to get chips fabricated. When you make circuits, there's a big wafer that they send through the line and put chips on it. The innovation was to take chip designs, pack them onto the wafer and share the cost of making that wafer among multiple designers. It was in fact the first version of electronic commerce that existed in the world. People would e-mail us their designs.

Q: How long ago was that?

A: It predates me, so more than 12 years ago. I'd say 15 to 20 years ago.

Q: Can you give an example of something ISI is working on now?

A: We're doing a lot of work in computer education, distance learning. Our particular slant is to use agent-based technology, so everyone can have an individual tutor. An "agent" (an artificial intelligence program) is watching what you're doing and tailoring lessons to your particular needs. It's going to be extremely powerful. I think education is in for a big revolution.

Q: What are the challenges with that program?

A: How do you write a course very quickly and efficiently? That's extremely, extremely expensive to do. The other thing we learned is that what people take from a professor, they won't take from a computer. When they look at a screen, they want Hollywood production values.

Q: What is your role at ISI?

A: I look at my role as the chief scientist and garbage collector. We have a certain amount of funds I can invest in new things and people. What are the new areas we need to get into that we're not into? What things should I fund? This is all run by government money, essentially. So we're constantly looking for contracts.

Q: How does it work when you invent something or create something and there is an opportunity commercially? Do you create your own company, license the technology to outside companies, or sell it outright to outside companies?

A: All of the above. This optical switch company got started because someone was going through the patent portfolio at USC. They came upon this and looked up the inventor and asked, "How would you like to start a company doing this?"

Q: What is ISI's relationship with and role in L.A.'s business community?

A: We've begun to spin out some companies. I think that's going to increase. I'm trying to encourage companies to stay here. I'm trying to see whether we can make contacts with venture capitalists who will fund things and nurture these companies. We're in a poor position to find the financial people and the CEOs. So we need support from the venture community. That's still relatively small here. It's growing, it's better, and we're able to do things. we couldn't do 10 years ago, but it's bootstrapping its way up. We'll be in great shape in a few years.

Q: How is the current boom in technology and tech companies affecting you?

A: One thing that's happening here as a result of all this VC stuff is that we're losing people like crazy. Our hope is to make it easy for them to (launch their companies) in this space. Hopefully these guys will make some amount of money, feel contented and then come back to the research game. At least that's my intent. Hopefully they'll be happy being rich but not super-rich.

Q: What programs here are brand new, or just recently began development?

A: There's something called "electrical elves." That's about six months old, but people are very excited about it. A simple example is, say you're trying to schedule a meeting with a group of people. Each person has an "agent" (an artificial intelligence program), and all the agents exchange information (via the Internet). They won't give out personal information, but if you need to reschedule a meeting, my agent will look at your agent's calendar and figure out optimal times. They'll also learn things based on prior experience. if there's someone who's a secondary person, the agent can decide to go ahead with the meeting with or without them. (The agents) negotiate with each other and learn enough about you to help you in your work.

Q: Would a user have to tell the agent what to do?

A: Sometimes yes, and sometimes it learns over time. That's the theory at least. Say it learns that you, when we have a luncheon, always like tuna-fish sandwiches and Diet Coke, if we have a luncheon, we'll bring in that meal, unless you do an override.

Another one that's just starting is something we call "digital amphitheater." Here the objective is to have a teleconference with hundreds of people (over the Internet). You have a little picture of everyone on your screen. Everyone can talk, and when they talk (their picture is enlarged to fill up the screen), and so on. We're figuring out the protocols. It uses all this Internet II bandwidth that people are talking about. Can I simulate a conference without having to leave my office? That's the objective of that.

Q: What are your passions beyond science?

A: You ever heard of (painter Jean-Michel) Basquiat? We were his first collectors and his largest collectors during his life. We knew him (from the New York art scene). Madonna was his girlfriend. We met Madonna before she was Madonna-in-big-lights. We knew (Andy) Warhol, William S. Burroughs, Allen Ginsberg.

Q: That's not average for a scientist.

A: (Laughs.) Somebody once said that I probably admire creativity as an art as well as in a scientific form. We bought art before we bought furniture. I was always interested in it. I don't know why.

Herb Schorr

Title: Executive director

Organization: Information Sciences Institute

Born: Jan. 20, 1936, New York

Education: Bachelor's degree, City University of New York; Ph.D., Princeton University.

Career Turning Point: Landing a summer job that introduced him to digital circuit design

Most Admired People: Charles Darwin, Albert Einstein and Pablo Picasso

Hobby: Collecting art

Personal: Married; three children