Cloning Silicon Valley

Published in IEEE Spectrum Magazine, May 2014

The speaker was talking about how New Jersey could become the next Silicon Valley.  I was skeptical, but the audience was enthusiastic.  "We could be Silicon Valley!" they were thinking.

Through the years I've been a number of places where they thought they could become Silicon Valley.  Some have done well on their own -- Boston, Research Triangle Park, Austin, and Cambridge, UK -- to name a few.  Many other places have had their hopes and aspirations come to little fruition.  Silicon Valley is still the only Silicon Valley.

While the speaker was lauding the potential of New Jersey, I was remembering back to a time in the 1960s when I had just started my career at Bell Labs.  Some executives had started an effort to make New Jersey into a Silicon Valley.  They had put together a consortium of New Jersey research organizations and hired Fred Terman, the Stanford dean given credit for creating Silicon Valley, to do something similar for New Jersey.

I have vivid memories of attending meetings where Terman outlined his vision for a New Jersey Silicon Valley.  The state had the greatest concentration of engineers and scientists in the country, but he said that what New Jersey lacked was a Stanford.  None of the existing universities had the necessary culture of engineering innovation.  But we could create a Stanford!  He proposed a new graduate university, making use of all the talent the state already had.  I was thrilled.  "We could be Silicon Valley!" I was thinking.

Terman said that the new university would have immediate credibility, and that universities were like cathedrals in small towns; when seen from a distance only the high spires were apparent, and New Jersey had those high spires in its famous engineers and scientists.

Needless to say, it never happened.  The consortium fell apart and there was no new university.  However, I seriously doubt that it would have led to a Silicon Valley.  New Jersey R&D was then a set of non-communicating islands, of which the largest was Bell Labs.  The geography kept engineers apart, and there were no "watering holes" where they met up.  The eastern culture favored academic publication instead of entrepreneurship.  And though I believe New Jersey's attractiveness as a place to live is underappreciated, the San Francisco Bay area is justly famed.  Moreover, New Jersey's proximity to New York City is both a blessing and a curse.  All roads lead to Manhattan, which is like a giant black hole sucking matter from its surroundings.

Even if an area has all the necessary ingredients in its environment to duplicate Silicon Valley, it would be unlikely to happen.  I think of it as a tipping point phenomenon -- winner take all.  Silicon Valley becomes the place to be, and it's game over.

It has now been about a half century since Terman tried to clone Silicon Valley in New Jersey.  During that time many other unlikely places have aspired to create their own Valley, and I still hear of ongoing attempts.  Perhaps it's time to find another model.  Some people have suggested a distributed model where each geographic area focuses on a particular area of competence.  Such a horizontal segmentation already exists to some degree with design, fabrication, software, and assembly taking place among a global community.

Still, there is some magic in having companies clustered face-to-face across the street.  All these years, and I'm still envious.