Leonard Kleinrock, a computer science professor, sat in his office in the basement of Boelter Hall, working late. It was past 10 p.m. on Oct. 29, 1969.
Down the hall one of his graduate students, Charley Kline, was also still up. Kline said he preferred working at night. During the day, the noise of the other researchers and the massive computer fans from the device in front of him and beneath the floor made focusing difficult.
A few minutes later, Kleinrock dropped by to check on Kline.
Two months earlier, Bolt, Beranek and Newman, a computer company, had airlifted in the world’s first internet router, built to the specifications provided by Kleinrock’s team. Stanford had received one at the beginning of October. Kline would be attempting to log onto the computer of Bill Duvall, a Stanford Research Institute computer scientist, using a simple computer command: typing “LOGIN” onto the welcome screen.
Kline had Duvall on the telephone, using the technology they intended to displace to see if their experiment worked, Kleinrock said.
Kline typed the first letter, “L.” Duvall received it.
Kline typed the “O.” Duvall received it.
Kline typed the “G,” and Duvall’s computer crashed.
“Oh, I didn’t get it, something seems to be wrong," Duvall said. "Let me call you back.”
It was just another step in the right direction of the aim of their research, Kline and Kleinrock thought at the time. The cascading implications that would follow wouldn’t become clear to them for another decade.
“You might wonder what crashed,” said Kleinrock, considered one of the founders of the internet. “It wasn't our computer at UCLA, it wasn't the first switch. It wasn't the high speed line between those two switches. It wasn't the switch on the other end. It was the SRI computer.”
A new technological era
Kleinrock said he was born to be an engineer. By the time he was 6, he had built a crystal radio out of household objects.
He was fascinated with engineering. It was a form of magic, he said.
By the 1960s Kleinrock was at the Massachusetts Institute of Technology, approaching completion of his master’s degree in electrical engineering. At that point, he nearly left academia – he had a wife and a young child, and he was ready to start work, Kleinrock said. It took his professors badgering him repeatedly for him to agree, he said.
“They really twisted my arm until I did,” Kleinrock said.
So if he was going to get his doctorate, it was going to be something more impactful and creative. His classmates were all working on small concepts of theory or extensions to current technology, Kleinrock said. Small, incremental pieces of progress, he called them.
Kleinrock wanted to work on a new system of transmitting data. He worked in MIT’s computer lab. He thought there should be a way to connect the computers to each other – to let them talk to each other.
At the time, the most advanced communications technology was the telephone. Whenever two people make a phone call, the line is open and being used until they hang up, said George Varghese, a UCLA computer science professor. It was an inefficient system because only one person speaks, but two people take up bandwidth.
Instead, Kleinrock said, information could be passed along circuits in small packages. These packages would shoot small bits of data from one computer to the next.
Kleinrock wasn’t the only person to develop the theory. Paul Baran, at RAND Corporation, and Donald Davies, a U.K. computer scientist, each reached the concept of packaging data at around the same time.
Kleinrock published his dissertation, which proved the mathematical theory behind packaging data, in 1959. In 1963, he took a job at UCLA, moving into the office on the third floor of Boelter Hall that he still uses today.
Kleinrock wanted to work on new, exciting research. The federal government was looking for exactly that.
In 1957, a few years prior to Kleinrock’s work, Sputnik launched and the space race began. By early 1958 President Dwight Eisenhower created the Advanced Research Projects Agency Network, or ARPANET.
ARPANET was a military-funded program, although its researchers would not work on technology for military use. ARPANET offered high-risk, high-reward contracts to researchers to essentially do what they chose with, Kleinrock said.
“That kind of flexibility led to some wonderful results across the country,” Kleinrock said.
Because most of the money funding research came from the government, provided freely to universities and researchers, scientific innovation was driven by curiosity, not profit, he said.
“In the '50s and '60s, there were things that people hoped can be done, but they were considered such huge projects (that) no company would consider doing them on their own, like the space program, like the ARPANET,” Kline said. “On the other hand, today, these are companies like SpaceX and we can go to the moon, we can go to Mars and can raise the money to do it.”
Nowadays, money is more difficult to come by, Kleinrock said, hindering innovation in the public sector – although the feeling of the era may be coming back, he added. But for a long time the scope of research in recent years has been more narrow, safer and more competitive, he thinks.
“It went pretty poorly for a long time,” Kleinrock said. “And you have far more faculty members competing for the small amount of money for the problems get smaller and narrower and highly competitive, as opposed to collaborative and cooperative. And so you're getting enormous amounts of papers published for little tiny results, which is exactly what I did not want to do with my dissertation.”
To facilitate collaboration and increase the pace of its various projects, ARPA decided in 1957 that it needed a way to allow researchers for its projects across the country to connect with each other. It needed a network. It developed a new project called ARPANET.
The head of the ARPANET, Lawrence Roberts, was a former fellow student of MIT. Because Kleinrock had the expertise, UCLA was chosen to be the first node to transmit data across a new computer network.
The first router and the first message
Kleinrock built a team of 20 to 30 people. They developed the specifications for a router that could connect with a computer, and ARPANET held a bidding war to choose a builder – AT&T and IBM both declined to participate, claiming it would never work.
Only 12 companies put in a bid. In the end, a contract to build four routers was awarded to the company Bolt, Beranek and Newman. To build those routers they altered minicomputers, creating what they called at the time Interface Messaging Processors, or IMPs.
During Labor Day weekend in August 1969, BBN airlifted its first router to UCLA, a few days early, to great general excitement, Kleinrock said. As administrators, company representatives, students and researchers crowded into the lab, one of Kleinrock’s students successfully connected the computer to the IMP.
On the other hand, when Stanford connected a computer to its own router a few months later, and Charley Kline transmitted data to Bill Duvall across the internet for the first time, there was nobody watching but Kline, Duvall and Kleinrock.
At the time, Kline considered it just another step in a long research process.
"Think of playing a video game and you finally got to the next level, you know,” Kline said. “You felt a little bit of excitement but nothing earthshaking, nothing like you had done something special.”
Duvall fixed his computer within half an hour, and Kline successfully completed the login. By Nov. 21, the first permanent link between two computers running on different programs was established between UCLA and Stanford.
The importance of the event went unrecognized for decades, even by UCLA. In the 1980s, UCLA attempted to throw out the original IMP to save space, so Kleinrock stole it, despite it being taller than an average person, and hid it in a small unused office.
Kleinrock asked the administration, his colleagues, even the mayor to recognize the importance of the accomplishment.
“(In LAX) there’s a big sign says, ‘Welcome to LA, home of the 1984 Olympics,’ Kleinrock said. “So (I said) how about a sign that says birthplace of the internet? Never happened, probably won't happen. But there was no appreciation.”
Meanwhile, the new internet’s capabilities were continuing to grow as more people entered the field.
UCLA received its IMP on Aug. 31, 1969. Stanford received its Oct. 1, UC Santa Barbara on Nov. 1 and the University of Utah on Dec. 1, Kline said. Each node was connected with each other node, forming the first computer network to be sustained between computers that ran on different programs. By Dec. 5, all four computers were part of a network.
From there, more universities and government agencies began to join the ARPANET. By 1974, the network had 62 connections.
Kleinrock’s team split up. Several took their own ARPANET money and built protocols to standardize the ways different programs would connect with each other, creating the first internet protocols.
The internet hadn’t been something planned, like the moonwalk or the telephone had been, Kleinrock said.
So it grew organically, individually, part by part, Varghese said.
We did not have a good message prepared, the way that Samuel Morse did, ‘What has God wroth?’” Kleinrock said. “Alexander Graham Bell, ‘Watson, I need you,’ or Neil Armstrong, ‘giant leap for mankind.’”
“(In the beginning) we were all well-behaved and friendly, I knew everybody,” Kleinrock said. “And so everybody behaved well, there was no profit motive. Nobody tried to patent, or gain access to intellectual property rights.”
By the 1980s, a darker side of the internet began to emerge, Varghese said. The original network programmers hadn’t put in place any protocols to authenticate users, so anybody could fake their identity on the internet. Security hadn’t been built in, Varghese said.
In 1982, Vinton Cerf, a former student of Kleinrock’s, and Bob Kahn developed the IP address system. Around then, corporations began to take a greater interest in the technology, Varghese said.
“As soon as it became commercial ... the bad guys are thinking, 'Hey, I can send a message pretending to be (anybody) right on the internet and nobody can tell because it's just a bunch of numbers,'" Varghese said. "And there was no attention to security, you know, built in, and retrofit security is very hard.”
Kleinrock correctly predicted in 1969 that the internet would see increasing application in the coming years. But he didn’t predict the rise and scope of its problems – fake news, cybercrime, radicalization – that would come along, too.
Some of it, at least, could have been solved earlier in the process, Kleinrock said.
“We could have turned on those (security) capabilities slowly and tried to prevent some of the abuses,” Kleinrock said. “Not all, but some.”
Kleinrock’s next project is coming up the day after the 50th anniversary of the internet – his new lab, just down the hall from the old one, is opening, he said. He said he and his graduate and undergraduate researchers plan to study everything to do with connectivity on the internet.
“The fact that it's just local should be inspiring to UCLA students, because any of them could have been part of some revolution like this,” Varghese said.
They hadn’t been prepared for the internet. But in just one of many happy accidents, the first word to be sent across the internet turned out to be fitting after all, Kleinrock said.
“We did not have a good message prepared, the way that Samuel Morse did, ‘What has God wroth?’” Kleinrock said. “Alexander Graham Bell, ‘Watson, I need you,’ or Neil Armstrong, ‘giant leap for mankind.’ ... But the message we ended up sending by accident was amazingly succinct, two letters, prophetic and powerful.”
“Lo,” Kleinrock said. Just like “lo and behold.”