Heading up the nonprofit SRI and its 1,400 employees is CEO Curt Carlson. He joined SRI in December, 1998 after 25 years with Sarnoff Corp., an SRI wholly owned subsidiary. As head of Ventures and Licensing at Sarnoff, he helped found more than 12 new companies.

How has SRI come through the current downturn many tech firms are experiencing right now?

We're growing and doing very well because we're not dependent on just one market. We do research around the world, license and continue to spin out ventures with real business models.

What's your take on what's happening in the Internet industry?

We were befuddled last year by what had been going on ... there were spinout companies without technology or even business models. We didn't know how to do that. When we do a spin out the company has to have significant barriers to entry and provide a benefit. It was hard to convince people that until they understood what kind of business they had, it's hard to launch a company successfully. In retrospect we look good. Our incubation model is the one left. The others are gone.

What are some of the hot Internet-related areas SRI is researching?

One is Geoweb, an architecture which lets you put tag every physical object in the world with an Internet address.

LATEST NEWS

UCSD Plans First Flash-Based Supercomputer
Digging Into N.Y.'s Antitrust Suit Against Intel
Analyst: Sony-Ericsson's Android Bid Is Late
Coupon Site Targets Black Friday, Cyber Monday
Microsoft Sites Up Big in Time Spent Online

Every object in the world? Please explain.

Today searching on the Net is done by key word, or .org and by business type. Those are powerful, but wouldn't it be nice, for example, to be able to say "Show me all the stores within a mile radius in Menlo Park" and you'd get a complete list? We own the technology that lets you do that, and it's extensible to anywhere on the planet.

How do you compile the list of objects? It's obviously beyond enormous.

The idea is that people and companies would self-register. A store, for example, would enter its location information. But even something mundane like telephone poles might be included because that might be something the power companies want to track. We have a working demo of Geoweb, we're incubating it and expect to turn it into a company.

Have most of the key Internet advances been made and now it's more a question of implementation?

We think we're just at the beginning. Navigating is one example. Today's search engines really don't work well. You type in your name, and you get back a few hundred hits, but not the article about the wireless Internet by Dave Needle that you were after. What's required is a technology we have that we're a leader in called natural language. You want the computer to understand what you're asking for in plain English and get back to you with the specific answer you're looking for.

Sounds like what AskJeeves does.

AskJeeves does what I'm describing, but it requires a lot of people in the back room researching answers. That model doesn't scale because the content on the Internet is already enormous and doubling every 4 to 6 months.

You're also doing a lot of research into Internet security.

Security at a human level hasn't even started yet. We can't even know we are who we say we are or if someone on the phone line is eavesdropping, which is so easy to do with today's technology. We're excited about the technology opportunity.

One thing we're doing is the development of a personal trusted network that solves the authentication issues and lets you have a secure, collaborative discussion.

From what I've read, the CIA and other government agencies have the kind of security you're talking about?

Yes, but it's very stove-piped, and violates the spirit of the Internet. You want to be able to safely broadcast around the world. We have a family of programs for security networks and tools to monitor the network so you can see what's going on, see if something looks suspicious and whether to let that entity stay connected.

We hear about this being the start of the post-PC era...

Well, we're excited about the wireless Internet, particularly in education. People are trying to put PCs in the classroom, but imagine if for a few hundred dollars you could give a rubberized, Palm-equivalent device, that could survive being dropped or spilled on, to a young student. When you move beyond the PC you can start thinking of using computers and the Internet in news ways. Kids can collaborate in smaller group projects and the teachers will find new ways to interact and communicate with them.

What do you make of the downturn in a broader, historical context?

We believe last year was an anomaly based on optimistic exuberance that was not justified. This year the industry is more pessimistic than it ought to be. VCs and others are preoccupied with mopping up the problems of last year's exuberance. It will probably take us 12 to 18 months to come out of it. But the Internet continues to grow. It's incredibly immature compared to what we dream about it becoming, yet it's still the most powerful communication media the world has ever invented.