“This Was Not a Money-Making Machine.” Internet Pioneer Leonard Kleinrock and the First Network Transcript

NOTE: The following has been edited for length and clarity.


 I’m Larry Magid. Several weeks ago, our producer Chris Le and I traveled to Los Angeles to speak with and tour the Laboratory of Professor Leonard Kleinrock, Distinguished Professor of Computer Science at UCLA. In the early 1960s, Kleinrock developed the Mathematical Theory of Data Networks, the technology underpinning the internet, while a graduate student at MIT.

He later accepted a faculty position at UCLA, and in 1969, the birth of the Internet occurred in his laboratory, when he directed the transmission of the first message to pass over the Internet between his lab and a computer about 400 miles north at Stanford Research Institute. So, Leonard Kleinrock, it’s good to meet you, and I’d like to, as we get into this interview, talk a lot about what you did right, and what you and other colleagues who helped create the internet might have missed, might not have thought about in terms of safety and privacy and security.

But before we get into that, let’s just start about how did you wind up at UCLA? I know just prior to that you had been at MIT’s Lincoln Lab. 


Lincoln Lab supported me for my full Graduate program, Master’s, PhD. Very beautiful fellowship and I had a position there when I was finished to do research at MIT Lincoln Lab. 

And I loved it. ​ But they said, look, we don’t want you to feel obligated to take a position here. Go look around outside. I looked around the labs, other places and you see, they thought I wanted a position. So I interviewed and they gave me an offer. Now here I had this offer from California; all the way across the country.

Larry: You were living where? 

Len: I was in Boston. 

Larry: In Boston. Okay. 

Len: And doing something I hadn’t done before. Whole family was back on the East Coast. 

Larry: Right. 

Len: And about half the salary. 

Larry: Oh, you had to take a pay cut. 

Len:  Oh, half. Yeah. Seven thousand dollars. That’s very cool for nine months, but it looked interesting. So I went to Lincoln Lab and I said, look, I’ve got this dilemma.

I really want to work for you, but there’s this job out there. Looks interesting. And the response was, “Try it. You don’t like it? Come back.” What a beautiful offer. It’s nice. Well, I’m still here.

Larry: And this was in 19—

Len: 63. 

Larry: So, yeah, I came here two years later, but I was in Political Science and Psychology at the time, so I didn’t cross paths with you. Uh, little did I know I would someday be in tech. 

So, there’s a lot I want to ask you. What was your relationship with Vint Cerf? We also talked to Vint and I’m curious, I know you guys worked together, but we didn’t get into that in our interview with him. 

Len: So, when I took on the ARPA contract, I put together a group of up to 40 people, including graduate students, programmers, staff, hardware experts, etc and Vint was among the students. He was a PhD student at the time in my software group. Mostly he was doing work for the ARPANET. 

Larry: Right. 

Len: Finding software for our host IMP interface. 

Larry: Right. 

Len: And the IMP we’re going to look at later, is this, the machine. 

Larry: Would you call it a router today? 

Len: Yes, you’d call it a router today. We called it a packet switch.

Larry: Right. 

Len: And BBN called it an Interface Message Processor. Hence the word IMP. They were playing around. 

Larry: That’s great. Tell me a little about the history of ARPANET and sort of when you came on board. 

Len: Sure. So the history goes back, I’ll take you way back, Sputnik, as you know, that was the source of all of this, right?

Larry: You know, for those who don’t know—there probably are people who don’t know—Sputnik was the first Russian satellite, first satellite that scared the bejesus out of our country because the Russians were ahead of us in space. 

Len: That’s exactly right. We were nowhere, right but they jumped ahead in an interesting way.

1957, 1958 was called the International Geophysical Year. Scientists across the planet were studying the planet, the atmosphere, the oceans, the rivers, the continents, and all the rest. And halfway through there, in October 1957, the Russians jumped ahead and they launched, as you say, Sputnik, started making this beeping sound as it circled the Earth.

First artificial Earth orbiting satellite. Beep! Beep! Beep! Drove everybody crazy. 

Larry: Could we hear it from Earth? 

Len: No. But it was a radio signal. The radio signal. Yeah. I remember I was around then. I don’t remember hearing it. 

Larry: Yeah. I didn’t think you could. That would have been amazing—

Len: —to have the power. 

Larry: Okay. So it made an electronic noise. 

Len: Yeah. And then President Eisenhower said, my gosh, they caught us with our pants down. 

Larry: Right. 

Len: We are no longer prime in technology and he said, that’ll never happen again.

Larry: Mm hmm. 

Len: So four months later, he formed ARPA, the Advanced Research Projects Agency, within the Department of Defense to do what?

To bring the United States back up in technology. So we started supporting research centers, universities, schools, all manner of technology. Just throwing money at it to bring up the capability. And among the other things they did was to support universities doing research across science. In 1962, ARPA created this Information Processing Techniques Office devoted to supporting computer research.

And the head of that was Licklider. Okay. And he came there with his wonderful vision of this man/computer symbiosis: put people and computers together and they can work together, each doing the thing they do best. And they started supporting computer science departments around the country. In fact, there were no computer science departments yet.

Larry: Really? 

Len: Yeah, they’re all Electrical Engineering. 

Larry: Okay. But all the early ENIAC, all that stuff was done by Electrical Engineering departments?

Len: Yes, yes. Not by computer science. 

Len: The earliest computer science departments came in the late ‘60s. We were one of the first. We came in ‘69. But it was slow going, but it was in electrical engineering largely, some mechanical engineering, some math, and a certain level of battles took place within the university bureaucratic structure, as “where will computer science go and who should go where?” But that’s a whole other story.

Larry: If you were a young person during the ‘60s, I mean, what were you guys thinking in terms of the social upheaval going on around the country. Now, that came later, probably, than when you first arrived, but it certainly came by the time you sent the first message in ‘68. What were you thinking in terms of what was going on in the country?

Len: So, understand, first of all, UCLA was not a hotbed—

Larry: Right.

Len: —of war resistance and…

Larry: Well, I went to Berkeley, yeah. I wanted to be part of that, yeah. 

Len: And, you know, it was an amazing place but, so, we didn’t get much of that here. There was not much sense of complaint or protest at all as engineers within an engineering department.

We were doing our thing. We were doing engineering. Ethics was not an issue. On the other hand— 

Larry: And at that point, nobody had any problem with the defense department. That didn’t come until the late ‘60s. 

Len: It didn’t. And there was no abuse of the technology as far as we could see. 

Larry: Right. 

Len: You know, who knows what was going on in other parts of the world.

But we were dedicated to solving problems. That was the challenge. That’s why we were engineers. To solve problems. So, there was not, no sense of ethics as a group. The culture that we grew up in the ‘60s, was one of ethics, we behaved well, trust, sharing, free, and fair. So, we didn’t—

Larry: And there was trust in government. I mean, my family were Democrats, Eisenhower was in office. It’s okay. Maybe they didn’t love all his policies, but there was a general sense of, you know, we were all in this together. We’d just been through a war. 

Len: Yes. 

Larry: And we were, you know, so it’s very different than today. 

Len: Well, we grew up in a wonderful generation. With plenty of opportunity. 

Larry: Right. 

Len: It’s not the case today. There’s a lot of threat and a lot of fear and concern. 

Larry: And we’ll get to that. But I want to, I didn’t mean to interrupt—

Len: ‘62. Licklider formed IPTO, right. Started supporting universities around the country in ‘64. Licklider stepped down and one of my classmates from MIT, Ivan Sutherland, stepped in as the head of the IPTO group. In fact, there were three of us very close together. We took our PhD exams simultaneously with the union of our committees. Claude Shannon, Marvin Minsky, names like this all together. So it’s quite an exam. 

Larry: Yeah. 

Len: Ivan Sutherland, Larry Roberts, and myself, we’re very close. In fact, Larry was an office mate of mine. 

So Ivan came in in 1964, and he knew of the work I had done, and we’ll get to that in a moment, as a PhD student on research of computer networks. And he recognized that we had three identical IBM computers here at UCLA. One in the medical school, one in the business school, and one for the campus in general.

And he said, let’s connect them together. Let’s make a little computer network. And it would have been easy. These identical machines, just a little triangle, and see how networking could work in a very simple environment. It never happened. 

Larry: It would have been a local area network in a sense, not that the term existed then, but you could have run a cable between these machines.

Len: Exactly. It would have happened. We would not have had to lease any lines from AT&T even. 

Larry: Right. 

Len: Well, it didn’t happen. The technology was there because of bureaucratic jealousies. People in medical school would say, what do you mean, you want to use my computer? For campus, we need 100 percent of the cycles, so the bureaucracy wouldn’t allow that.

Larry: And just, again, for folks that think of computers as something you’d get at Best Buy, these were giant machines, were millions. What did they cost back then? 

Len: Millions of dollars, yes. 

Larry: Millions of 1962 dollars. 

Len: Yes. 

Larry: Tens of millions of dollars today. 

Len: Yeah. But they were identical machines. It would have been easy.

Larry: Yeah. 

Len: Okay. Nevertheless, the idea of having a network, that idea was now pregnant in the ARPA community, in particular at IPTO where the computer work was being done. By the time…Ivan stepped down in 1966 and Bob Taylor came in Bob Taylor had already had four years of research going on. There were great centers of excellence around the country now. 

University of Utah had great graphics. University of Illinois had high performance computing, MIT, robotics, and AI. And so every time a new researcher came on, the researcher would say, “You want me to do research for ARPA? Fine, buy me a computer.” 

Larry: Yeah. 

Len: ARPA said, fine, “I’ll buy you a computer But look, see that stuff at Utah? I’d like that here. And Illinois, and MIT, and UCLA.” And ARPA said, “Wait a minute, we can’t buy everybody everything.” However, if you were in a network, and you wanted to do graphics, you could log on to the machine in Utah and run the graphics there. 

Larry: Right. 

Len: And hence, the idea of a computer network now became a real reality in ARPA.

So they brought in Larry Roberts to be Chief Scientist in 1966 at IPTO to manage this thing that was going to be called the ARPANET. 

Larry: So, in those days, if you were sitting at a computer terminal, before this, you were connected to the computer on your campus. 

Len: Exactly. 

Larry: And that was it. 

Len: That’s right. 

Larry: So, could you go across campus or did you actually have to be in the same room?

Len: Well, it depended on, typically those machines had this time sharing capability. 

Larry: Yeah. 

Len: So, you’d sit at a dumb terminal. 

Larry: Right. 

Len: And you’d have some kind of a connection to that machine. Some of it could be fairly long, a few hundred feet. 

Larry: Mm hmm. 

Len: Typically, they’re about 20 or 30 feet, and there were many users sharing the machine using the same idea of dynamic resource sharing.

You know, the very first computers were personal computers. 

Larry: Really?

Len: I sat at MIT as a research student in front of this huge expensive machine called the TX-2, one of the first transistorized machines. And I sit there debugging the program. I sit at this million dollar machine and I try a program and it wouldn’t work.

I said, “What’s wrong?” Scratching my head most of the time. It was being wasted. But in those days you could do that. And then the machines got more expensive, so they threw the user out of the computer room. 

Larry: Right. 

Len: And they did what you saw, submit your job in a set of punched cards. 

Larry: Right. 

Len: And basically it was called batch processing. Give it to an operator, they’ll run it for you, two days later you get your printed pile back. 

Larry: But you’re not happy. And when I did my dissertation, I had a—I didn’t have to pay it—but I got a $10,000 bill before they handed me my Doctorate degree, because I had wrapped up $10,000 worth of computer time from those IBM cards and other processing that I did.

And at that point, what I remember is when I was sitting at the University of Massachusetts Graduate Research Center on a terminal, I had a pretty quick connection. But when I was in my office across campus, it was much slower. I thought, stupidly, that the reason it was faster was because the CRT was at the University, it was at the Graduate Research Center, and my office had a DECwriter, which was like a typewriter, basically.

I later discovered something called baud rate. It had nothing to do with the speed of the terminal. It had to do with the fact that hard wiring versus telephone wiring. And I remember, I found that out very expensively when I actually bought an Apple II, used it as a terminal in my 300 baud modem, and it was still slow.

Len: So, that’s the answer. 300 baud modem. 

Larry: Yeah. But at that point, I think this is, you may know when the modem was invented—

Len: —but quite a bit later. It was a result of—the Carterfone case, where the government decided that AT&T had to allow users to apply other devices into their network. 

Larry: Right. It was a felony to hook up something to a phone number.

Len: Namely, an acoustic coupler. But the government stopped that. 

Larry: Yeah. 

Len: Allowed people to use couplers, so you could basically send beep, beep, beep through a network. 

Larry: Mm hmm. 

Len: That was phenomenal. Suddenly, every telephone in the world became a terminal.

Larry: Right. Talk about rapid deployment. 

Len: Yeah. 

Larry: It enabled quite a bit. But prior to this—so, fast forwarding to the IMP and the…You sent, what I gather was essentially what became email, right? The first message. 

Len: No, no. It was more of a message, not email.

Larry: Message, not email. But there was no email at this point.

Len: No email. That came in 1971. 

Larry: Yeah. 

Len: And we’re talking about 1969. 

Larry: Right. 

Len: In September we have one node and I’ve covered two nodes. Got a network. A network! One node is not a network, right? But two does. 

Larry: Yeah. 

Len: So now we could try out that basic idea. Sit at one location, log on through this network to another location, and run something there.

So the—all we want to do is test out, could we log in from here to there? That would have proven the test. So I have one of my graduate students, Charlie Klein, he was here one night. He made a connection with another programmer up at SRI, Stanford Research Institute, a guy named Bill Duvall. He’s in Menlo Park, 

Larry: About 400 miles north of here.

Len: That’s right. So it’s a distance, but we were connected with a high speed line, the first high speed line of the internet connected UCLA to SRI. Okay. At, never mind this 300 baud, we were going high speed. 50,000 bits per second. 

Larry: Oh, that is good. 

Len: Yeah. 

Larry: Today, you wouldn’t pay a nickel for that.

Len: Right. But in those days, that was broadband. So there it was. Let’s try the connection. So in order to do that, you have to type in “LOG” to try to log in and the remote machine, it’ll know what you’re doing. It’ll send the “IN” view. So all you have to do is to type “LOG.” So Charlie was ready to do this and uh, then we realized, wait a minute, now we’re not gonna see packets moving back.

How do we know what’s going on here? So we decided to get a telephone connection so they could talk to each other, you know, what do you see? 

Now look at the irony here. We’re using the telephone network. To prove a technology that’s going to eat the telephone network and of course it did. It got online.

Charlie typed the L. “Get the L?” Bill said, “Yep. Got the L.” And it printed on Charlie’s terminal. Type the O. “You get the O?” “Yep. Got the O.” Type the G, “You get the G?” Crash. Network went down. So, the first message on the internet ever was, lo. Lo. As in, lo and behold, I added that later. 

Larry: So that was also the first network crash, I guess, right?

Len: First network crash. First network crash. 

Larry: And who hasn’t experienced that in life, right? When your network goes down. 

Len: October 29th at 10:30 at night. 

Larry: So that was the first message and when would…you finally be able to actually get logged in? 

Len: Oh, within another half hour. 

Larry: Oh, so you were able to successfully…

Len: What happened was that well, what, what crashed?

It wasn’t our computer, it wasn’t our switch, the IMP, it wasn’t the high speed line, it wasn’t the switch at the other end, it was the SRI computer. Okay. So it wasn’t our fault. 

Larry: So that’s the other part about a network, is it takes two to function properly to be a network. 

Len: Yes. 

Larry: So I have to ask you, and this is really the crux, as you know, we’re ConnectSafely, we’re supported by companies like Google and Meta and others to promote internet safety and that includes civility, it includes security, it includes privacy, it includes lots of aspects of using the internet in ways that make us feel good and not bad. At what point did you come to the realization that this, that something bad could come of this? Now, I’m not, I’m not saying all bad—

Len: Of course. Plenty bad, though, 

Larry: Let’s not diminish it. Plenty bad and plenty good. When did it occur to you that we needed to be thinking about things like safety and security and other issues? 

Len: Much later, much later. You might ask about who were the visionaries who anticipated what we have today, and did they worry about that?

Okay. And I can show you one after the other, the answer’s no. I had a vision, I’ll show you a copy of it, put this in writing, before the switch arrived. Never discussed safety. Had good vision, totally missed social networking. 

Larry: Yeah. 

Len: Totally. But the idea of safety, early on we had hackers. Yes, we had crazy students doing things.

Larry: Mm hmm. 

Len: And they were a nuisance. In fact, we tried to hire one. He turned out to be terrible, by the way. 

Larry: Yeah. 

Len: The first time we really got hurt was in 1988. Robert Morris leached this virus, leaching out to quite a number of machines. And, uh, he claimed he didn’t let it loose, but he was a graduate student, he says his roommate let it loose.

Larry: Mm hmm. 

Len: We looked at it, we said, ouch! And then we said, oh well, just a hacker. What a mistake. It was a harbinger of what we have today. 

Larry: And spam. When did—I remember the famous case of the law firm that was sending out messages and people were very angry about that. 

Len: Canter and Siegel, April 12th. 1994. It was the first broad-based spam. There were some earlier ones, some that had commercial aspects to it, some were innocent, but the really big one that hit most of the machines was on April 12th, 1994. 

Larry: Make you feel any better, in 1984 I wrote a book called The Electronic Link, and I went back and looked at it. There wasn’t a word in there about security or privacy or safety. 

If you could go back and advise your younger self, and you were how old at the time, in your 30s?  

Len: In 1969, I was 35. 

Larry: Yeah, so if 35 year old Leonard, if you knew what current age Leonard knows, what would you have done differently, if anything?

Len: There were two things we could have done, but to answer your question, we were, you could say, innocent, naive, but in an environment where these evils were not happening. 

Larry: Right. 

Len: We were well-behaved. Our community was a small community. 

Larry: Right. 

Len: Well, I knew most of the people on the internet. 

Larry: Right. 

Len: And nobody even filed a patent or tried to take ownership. This was not a money making machine. This was a scientific engineering challenge. 

Larry: Right. 

And that was exciting. Things were working. Wow, new technology. The good applications, we could foresee some of them, that’s not what inspired us. It was engineering challenges. But in answer to your question, what could we have done?

At least two things. One is to put in good file authentication, which means that when I send you a file, there’s a way to prove that what you got is what I sent, and it’s not been changed or monkeyed with or replaced. Second, good user authentication. That really is who I’m talking to over this line. You know, these things were not big on our minds at the time.

And today’s problems are very serious. Look, we’re facing not only sophisticated hackers. We’ve got nation states. 

Larry: Yes. 

We’ve got organized crime. We’ve got a wealth of misinformation and false information. Those things are big problems, and they worry me today. You know, I’m afraid, among other things, that the internet is starting to get Balkanized, whereby countries are putting rings around their network.

Larry: Right. 

So these rings can’t talk to each other. And if they don’t, you’ve lost the effect of the internet. It’s open, free, accessible to everything. Can’t get that so easily anymore. 

Larry: And you also even make worse the current state we’re in, where we all are living in information bubbles, information silos, which is already happening in the United States because of media consumption and social media consumption.

But, you know, the idea of the, what was it, I think it was Moynihan said we’re all entitled to our own opinions, but not our own facts. But if the internet isn’t allowing you to get facts, you know, where do we go from here? 

Len: So if you ask, what is the solution? Well, you can address the problem not necessarily with the solution, but to recognize who are the players here?

Who are the stakeholders? There’s at least four I can identify. One, basically, big guys: the Googles, the manufacturers, the people who are giving us these services. They’re not acting responsible. They’re being pushed. I mean, only recently has Google started to ask you, what privacy protocol would you like applied to you?

They’re hardly asking that. Secondly, the government. What should be the role of government? Well, at a high level, the role of the government should be simply to create a forum whereby these stakeholders can talk. In some cases, there are things that the government needs to do because it’s universal. Third are the scientists.

We have to provide solutions, the kinds of things you’re asking of us, to help mitigate these problems. But the fourth group is the one I really complain about. That’s the user community. We’re not vocal enough. We’re not complaining. We’re not saying what our complaints are. We’ll, we’ll fetch a little bit, but as a group, we don’t have any power.

I mean, you should be able to say, no, I will not accept that privacy policy. And here is what I want. People are either shy or uninterested. 

Larry: So getting back to where you were and where we are today. So there are young Leonard Kleinrocks today who are getting jobs at universities. There are entrepreneurs, there are founders, kind of like you.

They’re motivated by wanting to do great things, yet they continue and create products which have unintended negative consequences. Today, you would think we would have learned something, right? 

Len: It’s a great question. I was going to answer a different question, which is what motivates those young people?

It goes back to the first spam message. Up until that first spam message, the ARPANET, the internet, NSFNET, was serving engineers, scientists, and not commercial interests. 

Larry:  Right. Right. You weren’t trying to try to become a billionaire. 

Len: But that happened in 1994. It took a sharp turn and said, “oh, this is a money making machine.”

And a lot of the effort, and now the young scientists are moving in that direction to create something that’ll generate money, suck money out of the consumer or the industry. And so a lot of the motivation is not to do great science, but to do a great product, which will in fact generate income. And that’s a different motivation than were the people I—we—love working with.

We are solving hard problems, good scientific stuff, you know, aha, exciting. You know, the thing that drove many of us into science, because we loved it, and that’s the way we feel as scientists. When suddenly the public got involved, the money makers, this became a, uh, a gossip chamber, became a money making machine, that’s what the internet became.

I wrote a not bad article: I pointed out we had this magnificent thing called the internet and suddenly it ended up selling detergent. What are the alternatives now? People are basically seduced into using them. These devices, and they want them, and they’ll pay a big price. They’ll avoid, they’ll give up their privacy.

By the way, privacy, forget about it. Game over. You gave it up a long time ago, when you put your address in an old telephone book. 

Larry: That’s right. 

When you carry around that phone, you carry around a credit card. 

Larry: Oh, credit card, yeah. Absolutely. 

So, we voluntarily give it up, but is there some limit? And those limits have to be described by that characteristic I gave you before.

We’ll give you this, but not that. 

Larry: And Europeans have a lot of rules about that. 

Len: Yes. 

Larry: Companies have to obey those rules. They operate in Europe, and they’re better than we are.

Len: Yeah. 

Larry: Far better. There’s a lot of worry about kids today, to quote a lyric from a song in “Bye Bye Birdie,” which was recorded many, many years ago.

But, you know, what about young people? Uh, should we be putting up barriers? Should we be encouraging their use of social media? There’s some good, but there are a lot of problems. 

Len: So you raise a good question, because we’re talking about kids. Kids these days have grown up in ways different than you and I did, and they live on that machine.

But they live at the social network level, not at the technology level. But many of them basically are strongly ethical, and they’ll complain about social ills. And how do you harness that capability, that power that they have? And they’re beginning to express themselves and maybe that’s the focus. Forget about old fogies like me.

But, you know, we went through a period when it wasn’t a concern, then it became a concern, and we don’t know what to do with it. These kids are living with it, in the concern and, I, they, some things they’re happy to allow. They’re not so much worried about privacy, but they’re worried about climate, and war, and starvation, and equity, and diversity, and all the rest.

So, how do we harness that power? I think that’s an approach, not a solution, but it’s a way to think about it. 

Larry: And that brings up a question. Here we’re, talking to you in your office at UCLA. I don’t know whether you have undergraduate students or graduate students, but obviously you’re around younger people.

Len: Yes. 

Larry: What are you learning from your students?

Len: We have an undergraduate program that we support, thanks to one of our Sloan alumni, which gives prizes to about a dozen undergraduates who come up with ideas about which they’re passionate. And this is across the university. The idea was that I recognized a long time ago, the undergraduate population are brilliant. There’s enormous talent there that we’re not tapping. 

And these kids, some of them have these ideas, they really want to do something, some project. They don’t have money to do it, and they don’t have time, they don’t have a laboratory, they don’t have a professor. And they gotta go work in the cafeteria. So the ideas die.

We wanted to capture some of those and basically encourage them. So they send in a proposal. “Here’s what I want to do.” Meanwhile, we select, say, a dozen mentors. Someone like you, for example. And it’s been going on for eight years now and it’s very successful. And to answer your question, the ideas are all over the place.

Just have to have something to do with the internet: an ethical issue, a social issue. You know, a technical issue, etc. And the idea is just wonderful to see, and they’re all over the place, and many of them are looking to do good, to deal with inequities, deal with social ills, deal with evil issues, as well as technology issues.

And so yes, in the answer to your question, the undergraduate population is, they’re very idealistic in many ways, and ready to take on hard problems. Sometimes naively, sometimes brilliantly. They’re really smart, and they’re worldly and idealistic at the same time. So does that give you hope? 

Larry: Very much so.

Len: Very much so. Now, in terms of the graduate students, I’m going to take you back to when ARPA started funding researchers. I mean, how did ARPA decide who to fund in those early days, in the ‘62s and ‘63s and ‘64s? Who were they giving money to? Well, they’d go to someone that they knew, like Marvin Minsky.

They’d say, Marvin, we know you’re a great scientist. Here’s a pile of money. Shoot for the moon. Go do something brilliant again. Failure’s okay. We’re not going to watch you. You figure out what you want to do. Go to another researcher, okay? Go do it. So it was not a democratic way to give money. It was not a competition.

It was very undemocratic, but it worked. It was a lot of money, and what would a Marvin Minsky say, or myself, do with the money? We don’t need it because we’re being paid by the university. We support graduate students. Same way. We don’t tell them what to do. We say, look, here’s the kind of problem. Go figure it out.

Talk to your colleagues. Talk to your classmates. You want some help? Maybe we’ll help you, but we’re not going to force you, we’re not going to say, do this in this time frame, don’t fail, and make sure, blah blah blah. That was the environment back then. That was part of the culture I tried to describe, but I didn’t get to that point.

You know, the honesty, it was the challenge. Then it started changing. The Mansfield Amendment came in, and you had to have a military application, you had to have a competition, and so the quality of that interaction went down and it went down pretty deep. It was one time—I won’t say who it was —came to me from ARPA and said, “Len, look, we want you to solve the following problem: We want you to provide two bits per second reliably out of the foxhole. Here’s $50,000, do it in six months. Don’t fail and we’re gonna watch you.”

I said, “Are you kidding?” I say, “Instead of asking me to build an internet, you’re asking me to get two bits out of the foxhole and you’re gonna force me to do it in six months?” The whole quality change what happened for a period was the funding agencies were given small amounts of money with far more people competing for it on smaller problems.

And the faculty were accepting that because there was no other source of money. And those faculty were showing their students: this is the way we do research. Little, constrained, don’t fail, blah blah blah. And micromanaging apparently. Micromanaging, very bad idea. 

Larry: Yeah. 

Len: Now who are those graduate students? They are the faculty of the next generation. So that propagated for a while. Hopefully, and happily, we’re coming out of that some now. 

Larry: Well, but in the corporate world that happens as well. I mean, IBM, and they still do great research, but used to do fantastic research. Microsoft, which is dedicating hundreds of millions of dollars to open ended research, and many of those programs have been dramatically scaled back.

Len: Exactly, they’re gone. Where’s Bell Labs, for example? 

Larry: Right, exactly. 

Len: Gone. IBM Yorktown, basically gone. And, and so where did those researchers go? They compete for faculty positions. And with the funding, there’s less funding, more people competing, and so that’s where we get to that small problem. So, this is a wonderful discussion, Larry, with just a lot of bad stuff, and hopefully there’s some—

Larry: Yeah, and we’ve already talked about the positive stuff, and you’ve already said that you’re very optimistic about the future. 

Len: Yes. 

Larry: So, this is the end of part one of my long and really interesting interview with Leonard Kleinrock. After this point in the conversation, we took a tour of his laboratory and later we picked up and we’ll get to that in the next episode.