Pioneering the Digital Frontier: A Conversation with Leonard Kleinrock, Computer Pioneer Award Recipient
For numerous decades, computer science pioneer, Leonard Kleinrock has been at the forefront of technological innovation. He stands as an illustrious figure who not only witnessed the development of what the Internet is today but also made vital contributions to the technology’s evolution. Dr. Kleinrock’s impact includes the development of the mathematical theory of data networks, shaping the landscape of modern networking, and underpinning the Internet as an MIT graduate student in 1962. At UCLA, where he now serves as the Distinguished Professor of Computer Science, his Host computer became the first node of the Internet in 1969.
Dr. Kleinrock has received eight honorary degrees, has published over 250 papers, authored six books, and has supervised the research for over 50 Ph.D. students. He’s received several honors, including the 2007 National Medal of Science, the highest honor for achievement in science bestowed by the President of the United States.
In recognition of his many achievements, we are proud to reward him with the 2024 IEEE Computer Society Computer Pioneer in Honor of the Women of ENIAC Award for, “the development of the mathematical theory of data networks, the technology underpinning the Internet.”
As someone who has not only witnessed but also actively contributed to the evolution of the Internet, what are your thoughts on its current iteration? Did you anticipate that the technology would be as ingrained in our lives as it is today, or would it evolve to what it is currently?
The Internet had more than 2 decades from 1969 to the early 1990s to curate and establish its “rules of engagement” and establish efficient technology and productive applications Toward the end of that period, as the network grew, we saw the dotcoms come online, the backbone channels escalate to gigabit speeds, and the World Wide Web with its convenient graphical user interface become a household presence. Once the interface was smooth enough and the audience large enough (by 1995, the Internet reached 50 million people worldwide), the commercial world recognized something that we had not foreseen: that the Internet could be used as a powerful shopping machine, a gossip chamber, an entertainment channel, and a social club. The Internet had suddenly become a money machine and the profit motive began to dominate its use. With monetization driving the Internet, the very nature of innovation changed. Averting risk has dominated the direction of technical progress. We have gone from a culture of creative collaboration to an arena of commercial competition, from agreement to antagonism, from moderation to megaphones of extremism.
Now, we are confronted with another massive technological development, namely AI and its latest manifestation as Large Language Models such as ChatGPT. This technology is on the threshold of converging with the Internet and will produce dramatic changes in the way that’s used (and misused). Although AI originated a decade before the Internet, its direction changed dramatically in the last few years and, unfortunately, this new LLM direction has not had the benefit of decades of curation and refinement. This leads to an uncertain future as to what this combination of technologies will bring even though both technologies have and can produce enormous benefits and value. I am optimistic about the future directions but recognize that it will require careful attention to how they develop and how their rules of engagement are designed and implemented.
Back in July 1969, I articulated a vision for the Internet that was published before the birth of the Internet. I described my vision as to what the network would look like in the future, and what would be a typical multi-computer network application. Importantly, at the end of the release, it contains a direct quote from me in which I articulated the following vision for the coming network, “As of now, computer networks are still in their infancy, but as they grow up and become more sophisticated, we will probably see the spread of ‘computer utilities,’ which, like present electric and telephone utilities, will service individual homes and offices across the country.”
I am pleasantly surprised at how the “computer utilities” comment anticipated the emergence of today’s Web-based IP services, at how the “electric and telephone utilities” comment anticipated the ability to plug in anywhere to an always-on and “invisible” network, and at how the “individual homes and offices” comment anticipated ubiquitous access. So, yes, I did see many aspects of today’s Internet. On the other hand, neither I, nor anyone else, envisioned a key ingredient of today’s Internet; specifically, I did not foresee the powerful community side (the social network side) of the Internet and its impact on every aspect of our society. It was not until email swept the network in 1972 that I realized the power of using the Internet to allow people to interact with people (i.e., social networks). Beyond social networking, I did not foresee the dark side emerging which now concerns so many of us.
Interestingly, there were earlier visionaries who could foresee what we now call the Internet. Notably among these was Nicola Tesla, who, in 1908, articulated an amazingly accurate prediction. So it is clear that the concept of an Internet was “in the air” and that vision had to wait for the technology to catch up to that vision. So, it is fair to say that the Internet was inevitable and would have emerged regardless if any of the “pioneers” had ever been born.
Honor your colleagues’ achievements. Nominate Someone for a Major Award Today!
Could you elaborate on the significance of the UCLA Host computer becoming the first node of the Internet in 1969, and how that moment influenced the trajectory of the Internet?
I had established the mathematical theory of computer networks in the early 1960’s. By the time ARPA decided they needed a computer network to support their research needs in the mid-1960s, I had a thriving laboratory working with me on the networking technology, so it was an obvious choice to select UCLA to be the first node and to serve as the Network Measurement Center to conduct tests, experimentation, and evaluation of the ARPANET. As a result, the origin period of the Internet was steeped in solid mathematical and engineering expertise within an academic environment.
Before the network was deployed, as good scientists, we knew how it should behave, and as good engineers, we knew that we didn’t know it all, so we knew we had to experiment, test, evaluate, and adapt the network implementation based on those tests and how they matched the theoretical predictions. Moreover, since this was a thriving academic environment; the graduate students were given significant freedom to create, explore and recommend ideas and technology. The culture was one where the faculty, the designers, the students, and the implementors interacted openly, freely, and creatively in a trusted, open, and shared environment. I coined a phrase to characterize that period, “Delegate authority to trusted parties.” This approach from the earliest days of the first node at UCLA did indeed influence the trajectory of the emerging Internet.
Moreover, the first experiment we conducted, namely, the transmission of the first message of the Internet from UCLA to SRI on October 29, 1969, was able to demonstrate the basic functionality of the network we were creating. That functionality was able to sit at a terminal connected to one computer at one location and log on to another differently configured computer at some other remote location and use its services as if one were a local user at that remote location. The experiment successfully demonstrated that basic functionality and set in motion what billions of users employ today, namely, to sit at your local computer (or smartphone) and reach out across the Internet to access what we now call a Web server to gain access to that server’s resources.
With the rapid advancements in technology and the Internet since its inception, how do you see the principles you developed in the 1960s continuing to shape the future of networking and communication?
The early culture of open, free, trusted, ethical, and shared working together has been the hallmark of the designers and developers of the Internet for decades. We established what we hoped were effective rules of engagement for the Internet. Moreover, we established a mechanism for engaging in the discourse using the effective mechanism of the RFC (Request for Comments) documents which is still in use. Further, it set in motion the approach to Internet development as exemplified by the graduate students’ creation of such groups as the Network Working Group (NWG). As Principal Investigators, we encouraged the students and developers to offer creative ideas and challenging goals. The encouragement of grassroots development from ‘below’ as opposed to the imposition of doctrines and standards from ‘above’ has proven to be highly effective.
Can you share some insights into your early days from childhood to your time at MIT, and how those experiences shaped your interest in computer science and networks?
As a poor child growing up in the streets of Manhattan, I enjoyed the open, challenging world of making things work in an unstructured environment. The independence that encouraged me has been a valuable strength.
My earliest engagement with technology began with a comic book! As a child in elementary school, I was reading a Superman comic when, in the centerfold, I found plans for building something called a crystal radio. This fascinated me because it looked like a fun and challenging project that would lead to a magical result, and I could build it for free with household parts I could scrounge around the apartment. No battery was needed and one would get music and radio programs out of it. I thought, “I want to try this.”
The first thing I needed was an empty toilet paper roll and that was easy enough to find at home. Then I needed some wire to wind around the toilet paper roll; this I found in a garbage can in the street. Next, I needed a razor blade which I found among my father’s bathroom supplies (I took a used one). In addition, I needed a piece of pencil lead to form the crystal for the crystal radio set. The next thing I needed was an earphone, which presented a small problem. However, I was aware that the public telephone booth in the candy store down the street had a handset from which I could extract the earphone. So, I went to the candy store, unscrewed the earpiece, and stole the earphone. The one remaining part I needed was something called a “variable capacitor”. For this, my mother and I took the subway down to Canal Street, at that time the Center for Radio Electronics was in New York City. We walked into the first shop we found and I walked up to the clerk and boldly asked to purchase a variable capacitor, whereupon the clerk replied with, “What size do you want?” This blew my cover, and I confessed that not only did I have no idea what size I needed, but also I had no idea what the part was for in the first place. After explaining why I wanted one, the friendly clerk sold me just what I needed for about a nickel. With all the parts now in hand, I went home, wired the whole thing up, and built the crystal radio set according to the plans. I then tuned the capacitor and adjusted the little crystal that I formed with the pencil lead and razor blade. Then music suddenly came in my ears. I was amazed to hear “free” music coming through the earphones – no batteries, no power, all free. This was magic! Where was this coming from? There seemed to be no energy going into the system. I just had to find out how it all worked. I was totally hooked!
From that point on, I would gather discarded broken radios, cannibalize the parts, and build new operational radios using minimal tools and almost no measuring instruments (I couldn’t afford them). I never became a HAM radio operator because I could not afford to purchase the relatively expensive “rig”.
The public school system in Manhattan at that time was truly excellent and I was fortunate to attend with really smart and dedicated kids. It was in Junior High School that I joined the Boy Scouts, and when I achieved the rank of Star Scout, my Scoutmaster challenged me to become the first Eagle Scout in our troop. This was a difficult challenge, but I accepted it, and when I did become the first Eagle Scout in the troop after considerable effort, I realized how rewarding and gratifying it was to set a goal that was essentially beyond my reach and to achieve it. Frankly, this gave me the confidence that I could indeed repeat that process in the future. I was fortunate enough to pass the entrance exam and be admitted to the Bronx High School of Science and Mathematics, which was at that time considered to be the best High School available.
I realized I could not afford to attend an out-of-state college, or, in fact, any college that required tuition, but I was fortunate to be admitted to the free City College of New York (CCNY). At that time CCNY was among the best colleges in the country serving an amazing collection of brilliant students. I was all set to begin my Freshman CCNY year in the Fall of 1951 as an Electrical Engineering student, but my father pointed out that not only would it be necessary not to pay tuition, but, due to his illness, I would have to get a full-time job and help support the family. This was a blow! Fortunately, he pointed me to his cousin who has a small industrial electronics firm in lower Manhattan designing and building photoelectric devices. I was offered a job and was happy to accept it since this was a great opportunity to learn about electrical engineering while attending night school at CCNY. Now what kind of folks go to night school to earn a bachelor’s degree in Electrical Engineering: crazies, drop-outs, poor dedicated students, and the GIs from World War II. This was quite an interesting mix. Of course, it was fraught with the danger of not succeeding and dropping out. Happily, I persevered, graduated in 5 ½ years (for a program that typically took the daytime EE students 4 ½ years), was the top EE student day and night, and was President of the evening session. All this while getting an amazing education at CCNY and gaining extremely valuable experience on the job as an assistant electrical engineer. The combination of learning the theoretical side of engineering at university and the practical side of engineering at work has been invaluable in my career.
Shortly before graduating, I heard that a researcher from MIT Lincoln Laboratory was coming to CCNY to describe a wonderful scholarship program at MIT for engineers. I attended the lecture, and he indicated that if one wanted an application to apply for the scholarship, to see the CCNY EE professor in the back of the room after the lecture. When I went to the professor, he said he did not recognize me and I told him that was because he was a day-time professor and I attended Evening Session. He said, “Evening Session – get the heck out of here!” I didn’t give up and contacted MIT directly and indeed won this terrific scholarship.
I then attended MIT for a Master’s Degree in EE under this MIT Lincoln Laboratory “Staff Associate” scholarship. Toward the end of my MS work, my MS Thesis supervisor suggested that I go for a PhD under the same scholarship program. I said I was not interested in a PhD and was happy to accept the job at MIT Lincoln Lab with my MS degree; by that time I was married and we had a child on the way. However, my supervisor “insisted” so I agreed to pursue a PhD, but I set two conditions for myself: (1) I wanted to work for the best professor that I knew at MIT (namely Claude Shannon) and, (2) I wanted to research a problem that would have impact and not merely be a problem that would be of no consequence. I did indeed work for Professor Shannon and generated a great problem of significance, namely to model, analyze, design, and extract the basic principles of how to create a computer network – a problem no one was working on. Interestingly, I observed that many of my classmates were working on extensions to work that their faculty had pioneered and I had decided not to follow that path, but rather to strike out on my own, namely not to “follow the pack.”
My experience as a PhD student at MIT and MIT Lincoln Laboratory was amazing. At MIT, my classmates shared their research ideas and projects, exchanged ideas, recommended approaches to each others’ research, and generally formed continuous conversations of shared ideas and individual breakthroughs. Our relationship with the remarkable faculty in communication theory and signal analysis was outstanding and students and faculty worked on research skillfully. It was an environment of discovery and progress. This set the tone for the research groups that I later formed at UCLA, and my PhD graduates, many of whom later became faculty at other universities and carried on that tradition. At MIT Lincoln Laboratory, there was a similar culture of idea generation, cooperation, and respect. The computer scientists who designed the earliest transistorized computers there in the 1950s were “complete” scientists in that they developed the architecture, created the design, formulated the instruction set, built the computers, invented the logic structure, tested the computers, wrote the compilers, generated the software and applications, etc.; in other words, they were complete computer engineers. One seldom sees that breadth of capability today. For me, it was an inspiration and reinforced my bringing together the practical and theoretical threads that make for an effective scientist/engineer.
When I was finished with my PhD at MIT, I was looking forward to taking the research position at MIT Lincoln Laboratory. My superiors at MIT Lincoln Laboratory wanted to be sure that I did not feel an “obligation” to work there because they had been so generous in supporting my MS and PhD program, so they suggested I look around to see what other opportunities were available at other research organizations. I did interview and got some terrific offers at places like Bell Telephone Research Labs in Murray Hill, NJ, but still wanted MIT Lincoln Lab. However, somehow UCLA thought I was looking for an academic position and ended up offering me a tenure-track position there. Now I had a dilemma since the faculty position sounded interesting (even though it was across the country, away from family, doing something I had not done, and at half the salary that MIT was offering me); I presented this dilemma to the folks at Lincoln Lab and, with amazing grace, they said: “Try it. If you don’t like it, come back to Lincoln Lab.”. What a wonderful reply. Well, I “tried” it and here I am more than 60 years later still at UCLA.
As a pioneer in computer science, what were some of the greatest challenges you faced when developing the mathematical theory of data networks in the 1960s?
Perhaps the first major challenge was to identify and define the problem to create a mathematical theory of computer networks and to understand and analyze these computer networks. No one had articulated or recognized the need to identify such a problem before.
When I formulated the mathematical problem, it became clear early on that finding an exact solution to that problem was intractable. It had all the problems of complex network flow problems as well as that of complex stochastic processes. So I had to find a way to “tame” the problem. This led me to introduce a critical assumption which, once introduced, allowed me to crack the problem and find a full solution.
That left me with the challenge to prove that the assumption I made was effective. To do so meant that I needed to write and run a complex simulation of computer networks and then show that the results with and without the assumption were sufficiently close. Writing that extensive simulation program on the MIT Lincoln Laboratory computer was a massive undertaking and if I could neither succeed in developing a running simulation nor prove that the results were sufficiently close, then I would not be able to complete my PhD. This was a huge challenge; fortunately, I succeeded on both scores.
Once I produced my results and graduated, I then took on the task of convincing industry that the ideas and principles I had proven mathematically were of value to them and that they should consider the development of computer networks. AT&T was then the largest network in the world and they totally dismissed my work, saying it would not work, and even if it did, they wanted nothing to do with it. It was not until ARPA recognized their need for a computer network that my work would finally be implemented.
Similar to what’s stated above, the industry in general often does not recognize the value of mathematical results for their product line. I have sought to change that view for many years, with some success.
You’ve supervised the research of over 50 Ph.D. students on various subjects. Can you share some memorable moments or achievements from mentoring these students throughout your career?
A principal mathematical tool for computer network research is “queueing theory”, an esoteric area of stochastic processes. It was always a bit of a challenge to convince PhD students to attend my courses on the subject, but once I showed them that queueing theory was the perfect tool for analyzing computer networks and more, they enthusiastically engaged.
In 1973, after one of my new PhD students and I attended an ARPA meeting where some ideas for packet radio technology were presented, we flew home to LA that evening and on that flight, we proved that the analysis that had been presented at that meeting was in error, and we obtained the correct solution to that problem. This then launched the beginning of that student’s PhD dissertation, and his results turned out to influence the entire field of wireless networks from then on. “From small beginnings, …”
I have always taught my students the importance of pursuing their curiosity and not only of solving a problem, but also of understanding the meaning of their results and how to extract general results from the specific problem they had solved. In turn, those who go into academia have been teaching the same approach to their students. Now this multi-generational collection of my “descendants” forms a small army of networking experts across the world.
I prepare very sophisticated problems for the final exams I give. Every so often, the problems are so “creative” that they lead me to publish them and then even lead to the subject of one of my student’s PhD research topics. This happened, for example in the subject of “Time Warp”, namely, L. Kleinrock, “On Distributed Systems Performance,” in Proceedings of the 7th ITC Specialist Seminar, Australia, September 1989.
Mentoring students goes beyond teaching them to do research but also involves showing them how to present their work effectively. I remember when I had a PhD student present one of our jointly-authored papers at a prestigious conference. To help him, I took him to the room in which he was to present a few hours before his session. The room was empty but was set up for the upcoming sessions. I showed him where to stand, how to use the overhead projector, how to adjust the microphone as he carried it back and forth as he cycled between the podium and the projector table, etc. When the time came for the actual session, he got up to speak, but we found that they had taped the microphone to the podium. So my student awkwardly spoke into the microphone at the podium, then would walk to the projector table to add a new slide and point to it, and then back to the podium, etc. This was a disaster and he had no idea how to manage the situation. I had no choice, so I got up out of the audience, went up to the podium, ripped the microphone from its tape, and handed it to my student who then was able to conduct the presentation sensibly and effectively.
As an aside, one of my PhD students is an amazing Texas Holdem poker player and shortly after he received his PhD, he won the year 2000 World Series of Poker.
You’re someone who has influenced generations of computer scientists. How do you actively engage with the community to inspire and mentor the next wave of innovators, beyond sponsoring PhD Students?
Almost a decade ago, I recognized the huge innovative potential that lies in undergraduate students at UCLA. Many undergraduates are passionate about an idea or project they would love to explore, but they typically have little access to mentors, funding, laboratory space, or faculty encouragement to explore their burning ideas. These students find themselves occupied with the daily work of classes, studying, and assignments, and may be employed part-time with jobs that have little benefit to their future careers or development; as a result, too often their innovative ideas are abandoned. I decided to create an interdisciplinary initiative (the Internet Research Initiative – IRI) which offers cash prizes to students from across the campus whose projects are selected through a competition. The selected students are offered a cash prize (funded by a former graduate student of mine), lab space, a mentor, and faculty support. The prize allows them to conduct research that through experiment, design, and analysis helps them put forward and implement their most revolutionary ideas at the intersection of Internet technology and society. I initiated this program in 2016 and it is now in its 8th year; it has more than 75 students who have passed through this program and has produced some amazing results.
In 1976, I published my second volume, “Queueing Systems, Volume II: Computer Applications” which was the first book to address the fledging Internet architecture and mathematics behind its design. I decided to “get the word out” on this technology to the research and industrial sectors as quickly as possible, and so, with my wife, formed a lecture company called Technology Transfer Institute (TTI). Our main goal was to have me conduct a 3-day seminar, “Computer Networks” across the country at three locations in the summer of 1976. It was a success, so we decided to bring in the best colleagues I knew to teach other computer technology seminars. TTI quickly grew, and so we hired people to administer the organization since I (as a Professor at UCLA) and my wife (a psychotherapist) did not want to leave our chosen professions. TTI has grown considerably for nearly half a century and continues to exist as it has evolved as a thriving educational institution serving to transfer knowledge and expertise to the practitioners and creators of digital technology.
I have created several research laboratories at UCLA over my career. In 2018, the latest research lab I formed is the UCLA Connection Lab to provide an environment that supports advanced research in technologies at the forefront of all things regarding connectivity. This and the earlier laboratories provide a welcome location for students and faculty to and visitors to gather and share ideas and lectures.
I give numerous keynote speeches and presentations to academia, industry and government, addressing the history and future of networking technologies (e.g., the Internet) and their impact. In my presentations, I point out how computers are the worst enemy of critical thinking and how important it is to explore the “how” and the “why” of the technologies we create work the way they do.
As an aside, the most significant location of the Internet is its birthplace, which is in my UCLA laboratory, room 3420 Boelter Hall. I carefully protected that location by preserving it as a historical room for the 55 years since the birth of the Internet. Moreover, the most important item in that room is the very first and original router of the Internet (we called it a “packet switch” then) and I managed to protect it for decades from being discarded as “junk” by literally hiding it in a room next to my office (and that was not trivial since the router is the size of a 7’ high telephone booth). Of the many dozens of similar routers of those early days, all but two remain (and ours is #1). My laboratory in 3420 Boelter Hall, along with its precious contents, was dedicated as an IEEE Milestone, one of IEEE’s most significant dedications. I have maintained this location for years, and insure that it is available to the visiting public on UCLA tours.
More About Leonard Kleinrock
Leonard Kleinrock is Distinguished Professor of Computer Science at UCLA. He is considered a father of the Internet, having developed the mathematical theory of data networks, the technology underpinning the Internet as an MIT graduate student in 1962. His UCLA Host computer became the first node of the Internet in September 1969 from which he directed the transmission of the first Internet message. Kleinrock received the 2007 National Medal of Science “For fundamental contributions to the mathematical theory of modern data networks …”, the highest honor for achievement in science bestowed by the President of the United States.
Leonard Kleinrock received his Ph.D. from MIT in 1963. He has served as Professor of Computer Science at UCLA since then, and was department Chairman from 1991-1995. He received a BEE degree (fully at night school while working daytime as an engineer) from CCNY in 1957 and an MS degree from MIT in 1959. He has received eight honorary degrees, has published over 250 papers, authored six books, and has supervised the research for over 50 Ph.D. students on a wide array of subjects, including packet switching networks, packet radio networks, local area networks, broadband networks, queueing theory, congestion control, gigabit networks, nomadic computing, intelligent software agents, performance evaluation, peer-to-peer networks and blockchain performance.
Dr. Kleinrock is a member of the National Academy of Engineering, the American Academy of Arts and Sciences, is an IEEE fellow, an ACM fellow, an INFORMS fellow, a CHM fellow, an IEC fellow, an inaugural member of the Internet Hall of Fame, a Guggenheim fellow, and an Eminent member of Eta Kappa Nu. Among his many honors, he is the recipient of the National Medal of Science, the Ericsson Prize, the NAE Draper Prize, the Marconi Prize, the Dan David Prize, the Okawa Prize, the BBVA Frontiers of Knowledge Award, the ORSA Lanchester Prize, the ACM SIGCOMM Award, the IEEE Leonard G. Abraham Prize Paper Award, the IEEE Harry M. Goode Award and the IEEE Alexander Graham Bell Medal.
