For Innovation: Be More Like Larry, and Less Like AT&T

For Innovation: Be More Like Larry, and Less Like AT&T

Before the dawn of the Internet, AT&T (America Telephone and Telegraph Company) were a mighty company and had created the world’s largest telephone network. The company had grown from the work at Bell Labs and, at its peak, employed over one million people. In today’s world, they would have a market size of around $112 billion (which would put them in the Top 10 companies in the world for revenue). But, they crumbled because of one thing: the packet-switched network.

Larry Roberts and ARPANET

In the mid-1960s, after much persuasion, Larry Roberts took on the role of designing a system which would allow mainframe computers within universities and defence agencies to intercommunicate — it sounds something that is quite simple now, but then it was an almost unthinkable task. If it could be created, it would allow research to be shared between academic groups and named ARPANET:

And, so, at the time, AT&T was all-powerful in the communications world, and little could advance without them. In fact, Larry Roberts had taken the advice from AT&T we would still possibly be using circuit-switched networks to intercommunicate, and you’d have multiple telephones in your house rather than multiple communication devices. But Larry had faith in his ideas, and went against those whose knowledge was stuck in the past.

Not circuit switched

For this, rather than following the advice from AT&T, he advanced the work of Paul Baran at the RAND Corporation and Donald Davies at NPL (Nation Physical Laboratory) in the development of a message/packet-switched network. Rather than creating a virtual circuit for the communication, they broke the data up into smaller packets of data, and which could travel independently over the network and find their own way to the destination. If some data packets were lost or had errors, they could be identified and resent.

Paul Baran had a military background, and in 1964, he proposed a method of communication that had no central control. and where if any of the intercommunication nodes were destroyed, we could still communicate — a distributed network. This went against the heavily centralised approach of circuit-switch networks, and which aggregated connections into centralised systems. Donald Davies, in the UK, had other ideas on this and envisioned a network which replaced circuit-switching with packet switching — and in doing so, he termed the word ‘packet’. It is interesting that Donald actually worked in a small team with the mighty Alan Turing.

It was then up to a little company called BBN to build ARPANET and break the power of the circuit-switched world. At the time, AT&T could see little potential in packet/message switching. Luckily, in the UK, we had BT, and who embraced these new innovations.

Don’t be like AT&T

AT&T just didn’t see how their telephone network could be replaced by anything else. Along with that, it was a threat to their mighty business and built around the innovations of Bell Labs. There were opportunities for them to help build ARPANET, but they declined and resisted any approaches. How could a little research project have a chance against their mighty communications infrastructure, and which was built on at least a decade of solid research and investment?

Conclusions

So, don’t let your company be like AT&T, and be ready to innovate and adopt new methods. If you can, read research papers whenever you can, and pick up on the latest methods before others get in before you. Larry was a great reader of research and picked up on the work of Paul Baran and Donald Davies through published work.

And my advice for government funding for innovation … think more Larry than AT&T. Get a great leader to drive things forward, and don’t rely on corporations to innovate anything.