Why is it called Silicon Valley?
A short walk from Stanford in 1957, eight men quit one company on the same day and started another. That single decision — and the corporate-fertility model it accidentally created — is the reason the global tech industry has a postcode in northern California and not somewhere else.
The bottleneck shift was sociological, not technical. Talent began moving between companies faster than companies could be founded, and the territory between Palo Alto and San Jose became the only place where that recombination loop ran at full speed.
Shockley brings the talent west
After the transistor work at Bell Labs, William Shockley wanted to commercialize the device himself. He left Bell Labs in 1955 and, partly because his mother lived in Palo Alto, set up Shockley Semiconductor Laboratory in Mountain View, California. He recruited some of the best young physicists and engineers in the country to work for him. The geographic accident — that he picked Palo Alto rather than New Jersey, Boston, or Chicago — is the entire reason the industry that followed clustered where it did.
Shockley was a brilliant physicist and a famously difficult manager. He was paranoid, micromanaged, played his hires off each other, and over the course of 1956-57 became increasingly fixated on a four-layer diode design that the rest of the team thought was a strategic dead end compared to the bipolar transistor.
By the summer of 1957, the most talented members of the team had decided he was unmanageable. They tried to get the company's investor, Beckman Instruments, to remove him. Beckman refused. So they did the only remaining thing — they all quit on the same day.
The Traitorous Eight
On September 18, 1957, eight men walked out of Shockley Semiconductor: Julius Blank, Victor Grinich, Jean Hoerni, Eugene Kleiner, Jay Last, Gordon Moore, Robert Noyce, and Sheldon Roberts. Shockley, who had grown up reading about loyalty to king and country, called them the Traitorous Eight in private correspondence. The label stuck and the men eventually adopted it themselves.
They needed funding. Kleiner wrote a letter to his father's investment banker in New York, a man named Arthur Rock. Rock arranged a meeting with Sherman Fairchild, an East Coast industrialist who controlled Fairchild Camera and Instrument and was looking for a way into the new semiconductor business. Fairchild fronted $1.5 million for the eight to start Fairchild Semiconductor, on the condition that he could buy the new company outright at a later date for $3 million — terms that look insulting today and that the eight would all come to regret.
Fairchild Semiconductor set up shop a few miles from Shockley's lab, on Charleston Road in Mountain View. By 1959 they had produced the silicon planar transistor, the manufacturing breakthrough that made the integrated circuit possible. By 1965 they were profitable, dominant, and unhappy.
Source: Christophe Lécuyer, *Making Silicon Valley* (MIT Press, 2006); Arthur Rock and Fairchild Semiconductor archived oral histories.
Why this clustered geographically
The technical work that followed could in principle have happened anywhere. What it could not have done anywhere is recombine. A semiconductor company in 1960 needed a critical mass of specialists — crystal growers, mask designers, lithography technicians, packaging engineers, applications engineers — and these specialists became more valuable the more often they could switch employers.
California had two structural advantages over every other plausible location. First, California's labor law had — and still has — an unusually strong prohibition on non-compete clauses (Business and Professions Code §16600). An engineer in California could leave a job and walk across the street to a competitor the next morning. The same engineer in Massachusetts or Texas would be enjoined for two years. Second, Stanford under Frederick Terman had explicitly set out to industrialize its surroundings, leasing university-owned land to electronics firms (the Stanford Industrial Park) and encouraging faculty to consult and to start companies. The university was acting as a venture-formation engine before that phrase existed.
These two conditions made the area uniquely permeable. A specialist who got tired of Fairchild could quit on a Friday and start at a new company on Monday, and the new company could be located within a fifteen-minute drive. The Valley was not built by one company succeeding. It was built by employees of one successful company leaving repeatedly, founding new companies, and recombining the talent pool on a six- to eighteen-month cycle.
The pattern that mattered
The thing to internalize is that the Traitorous Eight were not the founders of one company. They were the prototype for a process. In the decade that followed, alumni of Fairchild Semiconductor would leave to start an estimated 65 separate companies, including Intel (Noyce, Moore, and Andy Grove, 1968), AMD (Jerry Sanders and seven other ex-Fairchild employees, 1969), National Semiconductor (a related set of alumni, 1967), and venture capital firms including Kleiner Perkins (Eugene Kleiner) and the Sutter Hill funding nexus that Arthur Rock helped create.
These descendants — sometimes called the Fairchildren — went on to seed second- and third-generation companies of their own. Intel alumni founded later companies. AMD spawned competitors. Each successful exit became the seed funding for the next set of founders. The Valley's defining property is not a particular invention but this multi-generational compounding of talent and capital staying within a small geographic area.
Strategic read
The implication for any region trying to recreate Silicon Valley today is unflattering. You cannot copy the Valley by building research parks, university programs, or government VC funds. Those are necessary, but they are not the binding constraint. The binding constraint is the legal and cultural environment that lets talent leave employers freely, plus a critical mass dense enough that a person who quits today has a credible new job within a week.
Many regions have tried to recreate this. Boston's Route 128 had the technology and the universities but lost to California in the 1980s — Massachusetts non-compete enforcement is the most-cited culprit. Shenzhen has built a hardware-manufacturing equivalent. Tel Aviv has done it in defence-adjacent software. Bangalore has done it in services. Each successful cluster has its own version of the conditions: a labor law that lets engineers move, a few anchor employers whose alumni keep founding new things, and a density that lets a recombination cycle run faster than competing regions. The places that lack these conditions can produce world-class engineering but not world-class company formation.
For an AI-economy reader: the same dynamics now apply to AI labs, foundation model alumni, and inference-stack startups. The Bay Area is again the place where AI-research alumni from a small number of frontier labs are recombining into the next generation of companies. Watching where the alumni of OpenAI, Anthropic, Google DeepMind, and Meta AI go next is the equivalent of watching the Fairchildren spread out in 1967. The pattern is the same, and the geography is, by accident, the same too.