And I spent my summers in southern Mexico doing, a couple summers, doing anthropology, looking at people's adaptations to technology.

I went into law thinking that was where the rubber met the road in terms of applying understanding to that technology, to policy. And law school was fascinating. I knew I'd made a mistake about 5 minutes after I started work on my first day as a lawyer. It was like: Uh-oh. A classic case of getting the macro right but the micro wrong. You know. The big picture was correct, but the details were: this is a really miserable way to live.

Gemperlein: Depressing?

Saffo: Oh, I don't know. How would you feel if you'd just spent 4 years of your life working your brains out, taking bar exams in multiple states and found out that your chosen profession was hideous? Yeah. And I was living in New York on top of all that. That was not my happiest time.

But I also knew that I had lots to learn from the practice. In all honesty, I was a terrible lawyer. It was a struggle to go to work.

Gemperlein: How long were you a lawyer?

Saffo: About 5 years. First I worked for a large international law firm based out of New York and Tokyo and instantly was trying to do as much with computers as I could. This was 1980; the personal computer was barely in existence. But I was doing technology-related things from the start. And thinking more and more about it.

Talking about failures, my failure was I picked the wrong God-damn profession to be in. And I think all anybody who thinks about going into law, a priori, it's a rebuttable presumption, they've picked the wrong profession. I've made probably a (haha) pro bono side business out of counseling people contemplating law school trying to talk them out of it. But now, I'm 42, I cannot think of a single friend from law school who's happy in their job. Isn't that horrible?

You see stones from the Great Pyramid in Egypt and the names of the construction gangs are written on the bottoms of the stones. A pyramid wasn't built by slave labor; it was built by teams of workers who had extraordinary pride in their work.

I don't even have to go to other generations to think of people I'd like to talk to. I feel I have left unfinished conversations my whole life and there are a lot of people who aren't around who I wish I'd finished some conversations with.

David Packard is one. We don't have to reach back hundreds of years; we're surrounded by extraordinary people here in the valley today. And what we should all be doing is figuring out who should be going to dinner with who's around now and we could talk to them.

I think it's especially important for kids in school. I was real lucky because I had some mentors who reached out to me and had a big influence on my life.

And I lost a kite one day. I blew up less things than the average engineer in Silicon Valley. But I had my share of experiments. One was building a tetrahedral kite, designed and invented by Graham Bell. And I had decided to build the biggest tetrahedral kite you could design. It was about 5 feet high, and in a strong gust, it would lift me off the ground. And I then proceeded to put as much string on that kite as I possibly could.

And it was almost out of sight when finally the weight of the string, even though it's a hyper-stable structure, the kite finally tugged. And the string went past Don's window. The next morning, he showed up at the house in his old World War II Jeep with a long pole in back.

He had inferred what had happened. And we went looking for my kite. And that led to a friendship with an extraordinary man who really did have a big influence.

He has passed away. Some things happened to him that are very sad. His name's about to disappear. McDonnell-Douglas has been acquired by Boeing. He was a classic case of an extraordinary entrepreneur.

For every engineer who finds a unique slot and a unique idea, there are a hundred others who are replaceable parts. They get frustrated. Not able to go on. Get treated badly when companies grow and shrink in size.

Microprocessor-based analog devices. Things like MEMS, microelectromechanical systems, allow you to create analog sensors real easily. Anything that sends stuff out in this environment has to be analog.

So, we're about to see an explosion in cheap sensors, everything from video sensors to accelerometers and pressure-temperature sensors.

In the late 60s, it was learning machines. These were just horrible things. They were electromechanical. You'd kind of put in, well, sort of like an SAT exam. But a machine. Someone really thought this was a way to teach students.

The model that replaces it isn't clear yet. But I think I can say this without gagging __ I'm not sure -- that the idea where, instead of the person at the front of the room stuffing information into students like, you know, grain into a duck, headed to become pate on someone -- is instead a a wise companion and advisor.

I think information technology could be done right to make that possible. That it's a projection device where, you know, mentors can have influence and interaction with larger numbers of students.

The optimist in me says that's the way to go. The pessimist in me says that, you know, the nature of education in this country has got 2 big problems: It's terribly authoritarian, and it's terribly resource starved. Individual teachers are great. But, as a group, they're underwhelming.

They grab new technologies the way a drowning man grabs a life ring. And they try to solve short-term problems with it. And the result is that typically they create some new intellectual tiger cage to stuff students in.

The problem is that the education system is so screwed up right now that it's like trying to put a bucket of water on a burning building. My fear is that it's going to get worse before it gets better. It really has to collapse before people are really forced to do something with it. But, of course, the price of that would be terrible.

And I think, really, we have to start at the bottom of the system. The damage begins in kindergarten and first grade. If, by the time a kid gets to college. It's like me taking remedial calculus at Harvard. The very thought that a student could even get into college -- I mean, I'm grateful that Harvard let me in despite the fact that my calculus wasn't very good -- But that was something that should have been fixed much earlier.

The other thing is, the question about failures that we were talking about earlier, one always learns more from failures than one learns from success. Failure really stings when you're young and competitive. And if you're lucky, you figure out that you should love and embrace failure. And if you're not lucky, or you don't get the right advice, you learn to run away from failure, which causes you to stop taking risks. That was a lesson I think I learned by the time I was 20. Failing calculus was probably a big thing for me.

Doing badly in grade school was also an important thing for me. It had set my parents' expectations so low that they were so grateful that I went to college.

My poor father went to Yale, and I chose, of all places, to go to Harvard. And my parents -- both passed away recently -- and as I was going through their papers and things, I found that he could not bring himself to sign the checks to Harvard. My mother signed all the checks to Harvard.

And I think we're going to see some communities who say, We don't want to be connected. And others are going to say: we'll be connected, but we want a fence around us. In fact, there's a good example here in California. The two extremes face each other across a stretch of water. It's Stinson Beach and Bolinas. One city has gates on it, and the other one keeps tearing the sign down, pointing the way to the town. And they literally face each other across a little sand-spit of water.

So I think some communities are going to say, We really want the networks. And others are going to say: We don't want computer commuters. Look what's happened to Santa Fe. Aspen. And the segregation in society. Normal people don't live there. Real estate values shoot up. Aspen is the most dramatic example. The only people that can afford to live there or buy houses there are CEOs of Silicon Valley companies and Arab oil sheiks. You go to Aspen and go into a restaurant, buy a cup of coffee. The person that served you the coffee is probably doing a 40-minute to 1-hour commute from somewhere else in the mountains.

Technology can be democratizing, but it can also be very segregating.

The issue of haves and have-nots is really going to hit in the international realm. This is the stuff of real revolution. We have a bunch of companies in Silicon Valley who have research operations in the Santa Cruz district of Bombay. So these people, engineers in Bombay, or China, are going to work in the Santa Cruz district and, connecting by a high-speed link, do software design in the United States. And they go home at night into cozy houses, paid by salaries that are higher than normal rates would be otherwise. And outside their houses are people who, if they're lucky, they have a cardboard roof over their head. Bombay's a place where people literally sleep in the median strip of streets.

In the past, the rich have always been in a different place than the poor. And the poor haven't had it rubbed in their face that there are other people who are very rich. And now, in this shrinking globe of ours, impoverished Mexican peasants are watching reruns of Dallas. And people in Bombay are noticing that other people in Bombay are very, very rich because of their access to technology.

And when you put rich and poor close together like that, this is not a thing for domestic tranquility.