A Vision of Survival and Sharing
The story starts, as many from the era do, with the Cold War. The initial push for a new kind of network came from a simple, terrifying question: how could the U.S. maintain command and control after a nuclear attack? The existing telephone system, with its
centralized switches, was a major vulnerability. But the vision that truly shaped ARPANET wasn't just about military resilience. Thinkers like J.C.R. Licklider, a psychologist and computer scientist, imagined something far grander. As early as 1963, he wrote about an "Intergalactic Computer Network" — a way for researchers to share data and computing power effortlessly. It was a radical departure from the thinking of the time, where computers were isolated, room-sized behemoths. Robert Taylor, a later director at the Advanced Research Projects Agency (ARPA), was frustrated by having three separate terminals in his office to connect to three different computers. He saw the inefficiency and championed a unified network, securing the initial funding for what would become ARPANET.
The Phone Company's Billion-Dollar Blunder
Here's where history could have taken a sharp turn. Before building the network itself, ARPA did the logical thing: it offered the whole project to the undisputed king of communications, AT&T. In the early 1970s, ARPA program director Larry Roberts approached the telecom giant and essentially offered them the keys to the kingdom. The government would fund the research, and AT&T could take over, commercialize, and sell the service back to the government and others. After a serious review, AT&T's Bell Labs came back with a firm no. They concluded that the proposed system, based on a new technology called "packet switching," was incompatible with their own circuit-switched telephone network. They simply didn't believe it would work or be profitable. This decision, in hindsight a monumental corporate miscalculation, was a blessing for the future internet. An AT&T-controlled network would almost certainly have been a closed, proprietary, and heavily metered system—more like a cable TV subscription than the open platform that emerged.
A Packet of Genius
The technology AT&T dismissed is the very reason the internet works the way it does. Their network used "circuit switching," where a dedicated, open line is created between two points for the duration of a call. It's like having a private, continuous road from your house to a friend's. ARPANET, however, was built on "packet switching." This method, developed by minds like Paul Baran, Donald Davies, and Leonard Kleinrock, breaks data into small, labeled pieces, or packets. Each packet is sent out into the network to find its own way to the destination, where they are reassembled. It's more like the regular postal system: thousands of letters traveling independently through various sorting centers to reach their mailboxes. This design was revolutionary for two reasons. First, it was incredibly efficient, as multiple users' packets could share the same network lines. Second, it was inherently resilient. If one path was blocked or destroyed, the packets could simply be rerouted. There was no central point of failure, fulfilling the original military goal and creating a robust, decentralized foundation.
An Open, Collaborative Culture
Finally, the network's development process itself was a departure from the norm. Instead of a top-down corporate or military structure, ARPANET was built by a loose collaboration of graduate students, academics, and researchers at various universities. Wesley Clark suggested using separate, smaller computers called Interface Message Processors (IMPs) — the forerunners to modern routers — at each site to handle network traffic, a key decision that separated the network's function from the host computers themselves. The creation of essential protocols was managed by the Network Working Group, a collection of graduate students who documented their work in a series of notes famously titled "Request for Comments" (RFCs), a humble name for the documents that would define the internet's technical DNA. This open, academic, and problem-solving culture fostered the spirit of innovation and permissionless building that characterized the early internet, a spirit that would have been snuffed out under a more rigid, centrally controlled model.
