An internet founding father looks to the future

Leonard Kleinrock gravely concerned that funding for innovation may be drying up

Just as with any large project that grew in an organic, accretive way, the internet has no single founder.

The technology on which it was built relied on a large number of researchers based at various academic research institutes, building the internet precursor Arpanet and tackling the fundamental engineering problems posed by building computer networks.

Leonard Kleinrock is one of those early pioneers who made an invaluable contribution in the field, developing the mathematical theory behind data packet transmission during his PhD at MIT and while working at UCLA, where he is still a professor of computer science.

So significant was his input that he was one of the inaugural class inducted into the Internet Society's Hall of Fame, alongside the likes of Vint Cerf and Tim Berners-Lee.


And not only is he one of the most important figures in the history of the internet, he is also one of the most entertaining – at the Dublin Web Summit's Library Stage last week, he delivered a fascinating and entertaining overview of the development of the internet in the 1960s and 1970s, and offered his predictions on how it will continue to develop.

Revolutionary importance
As a young researcher, did he realise the revolutionary importance of what he was working on?

“When I did the research, the answer is no. I recognised it was solving a problem that needed solving. We’re surrounded by computers, and I saw there was no way they had to communicate. So I decided to attack that problem, an interesting problem full of challenges ... But what I missed was that my 99-year-old mother would be on the internet – I missed the social side. I thought it was about computers talking to each other, and people talking to computers. It’s not, it’s about people communicating with people.”

That eureka moment became clear with the development of email and newsgroups in the 1970s, when the computer network he had helped build slowly began to become a new communications medium.

Such a prediction would have been hard to make on the day, just over 44 years ago, when the first message was sent on the Arpanet – Kleinrock was supervising researcher Charley Kline’s attempt to make a connection from UCLA to Stanford with the command LOGIN.

It crashed after the first two characters, leaving the very first message on the internet at “LO”.

Not an auspicious start, and yet since then, those little seeds have grown into the global network and given rise to the technology industry that was being showcased at the Web Summit.

Kleinrock delivers these vignettes in a classic New York accent, despite his decades in California, and despite his 79 years he is not so much sprightly as effervescent.

The future, he feels, will be defined by nomadic computing, embedded tech in our environment, smart software agents and ubiquitous computing. "Dick Tracey will come to life again," he says, predicting the rise of wearable computing.

“We will have smart homes, and because chips are getting smaller and smaller, chips will be implanted in our bodies...The infrastructure is easy to predict – the internet will be a pervasive global nervous system. What is hard to predict is the applications and services built on that network.”

But he sounds a note of caution in one important respect – the environment he worked in was decidedly non-commercial, and it was a fertile scenario for lots of groundbreaking research. That innovation engine has been replaced to a large degree by the commercial model.

Funding pipeline
"There is a great danger that the funding pipeline is drying up," he says with evident concern. "Much of it did come from the federal government in the United States. But more than that it came with a culture behind it. The culture behind the early funding is what produced the golden era of Arpa, which produced not only the internet, but timesharing, graphics, high-performance computing, artificial intelligence. Why did all those great technologies be developed in a small period of time? The answer is that the approach to funding is very different to what we have now. The approach was they found an individual who was a good researcher, and then they threw money at him or her... We want high-risk, high-payoff approaches. It was a flexible, warm environment that allowed people to think big ideas."

Now however, we look to the technology industry and the start-up space for new innovations, and since the incentives are radically different, so will the outcomes.

"In terms of the funding profile, depending on industry to develop long-term, high-risk projects is a bad idea," he says. "Their DNA doesn't allow them to do that. Once they're on the stock market, they have to produce quarterly results. It's much like the innovators dilemma. Google might be an exception, they're trying innovative things all the time, not all of which succeed. You've got to get the universities, where the long-term, tough ideas are dealt with, with five, 10, 20-year horizons, not one-quarter, two-quarter horizons. Depending on industry is the wrong way – it's a good complement. Unless you have that mentality, you're not going to have the great breakthroughs. You're going to have incremental, add ons, and slow progress instead of these enormous leaps."

As a figure who was personally responsible for some of those enormous leaps, Kleinrock’s perspective is one we should heed if we want to encourage future breakthroughs on a similar scale to what he and his colleagues achieved.