The History of the Internet
The Network That Was Never Supposed to Survive a Nuclear War
The story you've heard about ARPANET — that it was designed to survive a nuclear attack — is wrong, and the truth is somehow more interesting.
The Idea
The myth runs deep: the internet was built by the US military as a communications network that could route around nuclear strikes. It's repeated in documentaries, textbooks, and dinner table conversations. But ARPANET's actual designers have consistently pushed back on this. Paul Baran, who did independently develop a concept called distributed networking with nuclear resilience in mind, worked at RAND Corporation — and his ideas were initially rejected by the military establishment before being quietly absorbed into the broader conversation. The ARPANET project, funded by DARPA and launched in 1969, had a more mundane and arguably more visionary goal: let expensive, scarce computers at different universities share resources with each other. What makes this reframing genuinely illuminating is what it reveals about how transformative technologies actually emerge. ARPANET wasn't built for crisis — it was built for efficiency. The four original nodes connected UCLA, Stanford Research Institute, UC Santa Barbara, and the University of Utah. Researchers wanted to run programs on distant machines and share data without physically transporting it. The nuclear resilience story got grafted onto the history later, partly because it made for a better origin myth, and partly because distributed networking does have that property as a side effect. The real intellectual achievement was packet switching — breaking data into small chunks, sending them independently across a network, and reassembling them at the destination. That idea, developed in parallel by Baran in the US and Donald Davies in the UK, is still the foundational logic of every data transmission happening right now.
In the World
On October 29, 1969, a UCLA student named Charley Kline sat at a terminal and attempted to log into the Stanford Research Institute's computer over the new ARPANET connection. The plan was to type the word 'login.' The system crashed after the second letter. The first message ever sent over what would become the internet was 'lo.' Accidental poetry. Within weeks the bugs were ironed out, and by December 1969 all four nodes were connected and functional. But what happened over the following years is where it gets philosophically interesting. The engineers building ARPANET kept discovering that what people actually wanted to use the network for wasn't remote computing — it was communication. Email, invented in 1971 by Ray Tomlinson (who also chose the @ symbol for addressing), quickly became the dominant use of ARPANET's bandwidth. Nobody planned for this. The social application ate the technical one. This pattern — a network built for one rational, measurable purpose being colonised almost immediately by human beings who want to talk to each other — recurred with every major communications technology that followed. The telephone was initially imagined as a broadcast medium for music and sermons. Early SMS was a technical afterthought. ARPANET's engineers watched their resource-sharing infrastructure become a message-passing system and largely shrugged, adapted, and kept building. That flexibility, more than any particular design decision, is what allowed the internet to become everything it became.
Why It Matters
There's a lesson buried in the ARPANET origin story that applies well beyond technology. We tend to reverse-engineer purpose onto successful things — to assume that because something works a certain way now, it must have been designed with that intention. ARPANET 'obviously' had to be about resilience. The internet 'obviously' had to be about information. But the actual history is messier and more generative: a small group of researchers solved a narrow, practical problem, built something flexible enough to be repurposed, and then watched the world take it somewhere they hadn't imagined. This should make you more skeptical of tidy origin stories — for technologies, institutions, even ideas. The things that last tend to be the ones with enough structural openness that people can bend them to new uses. And it might quietly shift how you think about your own projects: the question isn't always 'did this achieve what I intended?' but 'is it flexible enough that something better might emerge from it?'
A Question to Ponder
When something you built or started gets used in a way you never intended, is that a failure of planning — or a sign you built something genuinely useful?
Get a new one of these every morning.
Start learning with Thinkable