The secret history of Silicon Valley

This history is often overlooked or ignored – even (or perhaps especially) by the tech companies’ own workers. But it has profound implications for how that tech operates today and could be used in the future. Last year, for instance, Google’s involvement in Project Maven – which seeks to automate intelligence gathering and analysis for the Pentagon – led 3,000 employees to protest against their company’s work on ‘weaponising’ AI. You do wonder what these highly educated men and women thought Google had been working on all that time.

“In the last couple of decades, the industry has gone from solving problems in physical space to bigger issues of cybersecurity,” said Steve Blank – entrepreneur, academic and historian of Silicon Valley. “We’re now wiring the data world for our intelligence agencies.”

It’s easy to be seduced by conspiracy theories like the ‘Google Military Industrial Complex’ when looking at the complex relationship between Silicon Valley and the military. But it’s important that we try to understand this sometimes-stormy relationship in a more sophisticated way, as the global power balance starts to shift. Books such as Yasha Levine’s Surveillance Valley: The Secret Military History Of The Internet and Margaret O’Mara’s The Code: Silicon Valley and the Remaking of America have made a start.

Levine writes:

“Google is one of the wealthiest and most powerful corporations in the world, yet it presents itself as one of the good guys, a company on a mission to make the world a better place and a bulwark against corrupt and intrusive governments all around the globe. And yet I discovered that the company was already a fully-fledged military contractor…And Google isn’t alone.”

The tech giants may want to bury their defence department involvement, but war is in the DNA of the Valley. Take Palantir Technologies, the controversial, $10 billion data analysis firm, set up by PayPal co-founder Peter Thiel. It is said to wield as much real-world power as Google. Its first outside investment came from the CIA: its tools were developed for tracking insurgents in Iraq and Afghanistan – and are now used by the FBI, Homeland Security, and police forces across the United States. The New Orleans Police Department notoriously partnered with Palantir to pioneer “precrime”, or predictive policing.

And then it hit Wall Street. When the senior management of JPMorgan lost control of the group whose job it was to identify security threats inside the bank, Palantir’s software was parachuted in as a solution. A unit, led by a former U.S. Secret Service agent named Peter Cavicchia III, used it to aggregate, search and analyse employees’ communications, looking for incriminating evidence. In other words, Palantir’s software – backed by Cavicchia’s paranoia – did what it had been designed to do for the military: it started to hoover up emails, browser histories and GPS from company smart phones – and even digitally record phone conversations.

And it did so with alarming ease.

The project ended when the senior management realised that they were under suspicion as well.

Palantir’s ‘War On Terror’ technology has now found a wide range of uses – from tackling fraud and abuse, to understanding how people use data. The company’s clients include the World Food Programme, Airbus and the World Atomic Energy Authority.

The problem with all of this is that the dark side of the Valley has no physical, visible manifestation: there aren’t the chain link fences, security guards or use-of-lethal-force signs that you find at military installations such as Los Alamos, Skunk Works and Area 51. The original, small-scale nature of its work meant that it never had any oppressive physical presence.

In fact the Valley did have its secret facilities, according to Steve Blank, but these could be hidden away in a desert on another continent. (Blank describes working in a top-secret facility back in the 1970s whose technology was 30 years ahead of the commercial industry: “Nothing I would work on in the next 30 years was as bleeding edge.”)

We don’t see that. We see the gimmicky, giant-sized deck chairs and ping pong tables that the Valley uses to reinforce that hippy-libertarian founding myth. But the Californian orange orchards weren’t cut down in the race to build microchips – or to make room for giant deck chairs. They were felled in order to develop the microwave and radar technology which would defeat the Nazis’ air defence systems, and then to warn of Soviet missile launches and spy on their air defences. The Pentagon was the only body with the resources and the time to fund the kind of future-shaking research needed to win these hot and cold wars.

Stanford University became, in effect, the research laboratory for the American military-industrial complex. The now forgotten Fred Terman built up the 800-strong Electronic Warfare Laboratory at Harvard during the war – and, when the war ended, returned to Stanford to do the same again. It was routine for theses there to be classified.

The first IPO (initial public offering) out of Silicon Valley was in 1956 for a company called Varian, which sold microwave tubes for military applications. The first contracts for Fairchild Semiconductor were to help build the bomber and missiles that could lead to nuclear Armageddon – as well as take America to the moon. Lockheed Martin arrived in the Valley in 1956 to build missiles and then spy satellites – and quickly became the largest employer in the valley. The Valley was said to be then full of Soviet spies, as it is undoubtedly a target for Russian and Chinese  spies today.

Then, in 1957, the Soviets launched Sputnik and the US military panicked. The result was the Advanced Research Projects Administration – DARPA – set up to close the technology gap.

The Internet grew out of the Arpanet, funded by the transfer of a million dollars from a ballistic missile defence programme, which was, in turn, funded by DARPA. Its main aim was not surveillance, but to ensure that the US stayed ahead of its Soviet enemies in science and technology. The researchers behind it hoped that academics would work more effectively together if their computers were connected. The military did have a ‘secure capability’ on the Arpanet – which well have been used to transfer classified data acquired through surveillance – but there was no system of surveillance. Yet.

However, Levine pushes this argument further than most commentators have been willing to do. “The internet was hard wired to be a surveillance tool from the start,” he wrote in Surveillance Valley. “No matter what we use the network for today – dating, directions, encrypted chat, email, or just reading the news – it always had a dual nature rooted in intelligence gathering and war.”

There were even those in the intelligence services who dreamed of a “a sort of early warning radar for human societies, a networked computer system that watched for social and political threats and intercepted them”. It is hard not to hear the echo of this idea in the work of companies like Google and Palantir.

The Central Intelligence Agency set up its own DARPA-like organisation in 1999 “to ensure that the CIA remains at the cutting edge of information technology advances and capabilities”. Now known as In-Q-Tel, this fund invested in Keyhole, a satellite imagery company. That was sold to Google in 2005, where it became Google Earth. In-Q-Tel also invested in Palantir and helped it secure contracts to work with the US government on cybersecurity.

And then there’s Siri. The little voice that emanates from iPhones across the globe was developed from a project backed by SRI International, a non-profit research organisation with funding from DARPA. The goal was to integrate artificial intelligence into a virtual assistant that could learn and evolve by itself. “[It] was an extremely ambitious project, beyond what could be done commercially,” said Adam Cheyer, a Siri co-founder.

A look at the Pentagon’s latest National Defence Strategy identifies eight commercial technologies that the US military wants in on in the future, including advanced computing, big data analytics, artificial intelligence, and robotics – and we can see this agenda in Project Maven, perhaps the Pentagon’s biggest high-tech project to date.

The Algorithmic Warfare Cross-Function Team – to give Project Maven its proper name – uses AI to automate the analysis of huge amounts of drone data. The goal is to improve on their work of identifying threats and tracking enemy movements, by detecting anomalies that the human eye may miss. With the help of Amazon and Microsoft, it is already operating in least five secret locations in the Middle East and Africa.

To build on the success of Maven, the Defense Department has launched a $10 billion tender for an enormous cloud storage centre to store all this data. They have given it the code name JEDI (Joint Enterprise Defense Infrastructure) – and this is a contract that all the Valley wants a cut of.

“We’ve been told again and again that we are in the grips of a liberating technology, a tool that decentralises power, topples entrenched bureaucracies, and brings more democracy and equality to the world,” writes Levine.

“But spend time looking at the nitty-gritty business details of the internet and the story gets darker and less optimistic. If the Internet is a truly a revolutionary break from the past, why are companies like Google in bed with the cops and the spies?”

We would do well to remember that technology tested on the streets of Basra sooner-or-later finds it way onto our high streets. War is at the core of the tech that’s integral to modern life and risks propelling it to a darker future.

---