LOS ANGELES — The internet didn't start with a bang. It started with a crash, two letters typed into a terminal at UCLA, and a phone call to a lab 350 miles north in Menlo Park.
Related: THE TRAITOROUS EIGHT: HOW A 1957 WALKOUT IN MOUNTAIN VIEW BUILT SILICON VALLEY · The Silicon Valley Shortcut: Why Everyone in Tech Is Suddenly Whispering the Word "Distillation"
That October night in 1969, a graduate student at the University of California, Los Angeles sat down at a hulking computer terminal to log into a machine at the Stanford Research Institute, known today as SRI International. The plan was simple: type "LOGIN" one letter at a time and have each character confirmed by the receiving machine. The student typed "L." It arrived. He typed "O." It arrived. He typed "G," and the system on the other end crashed.
The first message ever sent across what would become the ARPANET, and eventually the internet, was "LO." As in, "Lo and behold." Nobody planned it that way. It just happened to be poetic.
Few tourists stop for it, but the moment is well-documented in university archives and widely cited in computing history. It happened before Apple existed, before "Silicon Valley" had settled into common use, and decades before anyone in Cupertino or Mountain View thought to build a campus shaped like a spaceship. The infrastructure of the modern digital economy has roots in California that predate the personal computer by years.
The ARPANET was funded by the Pentagon's Advanced Research Projects Agency and built to let research computers at different universities talk to one another over long distances. UCLA was chosen as the first node partly because a professor there, Leonard Kleinrock, had done foundational work on how data could be broken into packets and routed across a network. SRI, up the coast, became the second node. A second connection between the two campuses that same month is generally regarded as the first true host-to-host link on the network: the moment two machines actually talked rather than one simply calling and failing.
What followed was not an overnight sensation. The ARPANET grew node by node, campus by campus, largely unnoticed outside a small circle of researchers and government contractors. But the seeds of California's tech identity were already there: public research universities doing government-funded work, private labs experimenting on the side, and a culture, even then, of treating computers as tools for connection rather than just calculation.
SRI's role in this story runs deeper than a single network node. A year before the ARPANET connection, a researcher there named Douglas Engelbart staged what has since become known as "the Mother of All Demos," a 1968 presentation in San Francisco that showed off a working computer mouse, on-screen windows, hypertext links, and real-time collaborative editing. Ideas that wouldn't become standard consumer technology for another two decades. Engelbart's lab in Menlo Park was, in effect, previewing the graphical, networked computing that would define the industry built around it.
None of this happened in a garage. It happened in university computer labs and federally funded research institutes, staffed by professors and engineers who were, for the most part, not trying to build companies. The commercial instinct came later, layered on top of research infrastructure the state's universities had already built.
That distinction matters for understanding how California's tech economy actually formed. The valley's reputation rests heavily on entrepreneurship and venture capital, on founders who dropped out of Stanford or left stable jobs to chase an idea. But the ARPANET story is a reminder that a lot of the foundational work was done inside institutions that answered to federal grants and university administrations, not shareholders. UCLA and SRI weren't competing for market share. They were trying to prove that computers in different cities could exchange information reliably.
The network kept expanding through the 1970s, connecting more research institutions across the country and eventually overseas. The protocols developed to manage that traffic, refined over the following decade, became the technical backbone of what we now call the internet. By the time commercial internet service providers began signing up ordinary households in the 1990s, the underlying architecture had already been tested for 20 years on a system that started with two machines in California failing to complete a login.
Both of the original nodes still stand, after a fashion. UCLA's computer science department has preserved the room where the first message was sent, drawing visits from students and historians curious to see where the wiring lived before it had a name.