این جمله ای هست که در طی سالهای اخیر در تمام کشورها شنیده شده است:
"می خواهیم سیلیکون ولی خودمان را خلق کنیم."
“We want to create our own Silicon Valley.”
If there’s a single sentence I’ve heard in every country I’ve been to, it’s this one. Silicon Valley has been home to technology-driven innovation for a long time, but the 20-year period from 1994 to 2014 was something special.
People all over the world witnessed a spectacular level of innovation and wealth creation, all emerging from a small 30-mile long, 15-mile-wide strip of Northern California. Other states and countries have been attempting to build the “next Silicon Valley” for years now. At this point, there’s even a formula.
As Marc Andreessen writes:
The popular recipe for creating the “next” Silicon Valley goes something like this:
- Build a big, beautiful, fully equipped technology park;
- Mix in R&D labs and university centers;
- Provide incentives to attract scientists, firms and users;
- Interconnect the industry through consortia and specialized suppliers;
- Protect intellectual property and tech transfer;
- Establish a favorable business environment and regulations.
It happens all the time all over the world. And it never works.
When I’m asked, “What can we do to create our own Silicon Valley?” my response surprises many people: “You can’t,” I say. “It’s too late. Silicon Valley has a decades-long head start creating the perfect environment for creating Internet businesses. What you can do, though, is position your communities to compete and succeed in those areas of innovation that are still to come.”
The development of fields such as genomics, robotics, and cyber will all benefit from the interventions that Andreessen listed. But for cities or countries seeking to create the next hotbed for any of these fields, there are also broader factors to consider.
Building an innovation-rich place like Silicon Valley requires specific cultural and labor-market characteristics that can contradict both a society’s norms and the more controlling impulses of government leaders.
A critical factor: domain expertise
With the industries of the future — a topic I researched extensively as part of my new book — new avenues of opportunity for countries and people alike will hinge on domain expertise: deep knowledge about a single industry, which tends to concentrate in specific cities or regions. Detroit has domain expertise in cars, Paris has it in fashion, and Silicon Valley has it in Internet-based businesses.
Domain expertise for the industries of the future is still broadly distributed. To understand domain expertise, consider the following question: Why do a ridiculously high percentage of Internet companies still come out of Silicon Valley when massive investment is being made around the world to compete with it? Many factors are at play, but domain expertise is the most important.
For more than 20 years, the world’s best computer scientists have overwhelmingly been based in Silicon Valley. They could have been born anywhere, but they came to Silicon Valley for school (Stanford or Berkeley), employment (which creates a self-reinforcing cycle that concentrates talent), and investment (with the Valley offering far and away the most access to early-stage capital in the world). And they came to be included in a culture and community that placed the computer science engineer at the highest level of social status.
The Valley came to be not just any old industrial center, but a kind of beacon—a place that promised not just opportunity but a sense of belonging—and that continues to attract wave after wave of ambitious entrepreneurs. But nothing like that exists yet for the industries of the future, where the most interesting and important innovations are taking place with greater geographic spread than we see with Internet-based innovation.
There are early geographic leaders in each of the fields, but it is still far too early to describe any of these as the winners or losers in the competition to be home to the next generation of innovation. And what concentration there is today is not destined to be permanent.
In the current landscape, the most important work in the commercialization of genomics is clustered around universities where much of the original research and development took place. It is in and around Boston because of Harvard and MIT, Baltimore because of Johns Hopkins, and Silicon Valley because of Stanford and the Universities of California in San Francisco and Berkeley. Walking through the offices of these companies, one can’t help but notice how diverse the workforces are. European, Asian, African, and South American employees fill these companies and live in Boston, Baltimore, or California because they all studied at American universities.
The other major prong of genetics research is in China. Though it does not have a top university program in genetics, China has done an excellent job recruiting its citizens back home after they have studied abroad. As a result, Beijing is quickly becoming a center of domain expertise in genomics. In cyber, the most interesting companies are often based proximate to government, where domain expertise was developed inside the best law enforcement and intelligence communities, including Washington, D.C., Tel Aviv, London, and Moscow. Europe’s first cybersecurity accelerator, CyLon, was cofounded by two top foreign policy aides to British prime ministers. One of the world’s largest cybersecurity companies, Kaspersky Lab, is full of former Russian military and intelligence officers.
Israel has many of the best cybersecurity firms, founded by people who got their start in cyber in the Israeli Defense Forces, especially Unit 8200, Yehida Shmoneh-Matayim, the intelligence corps focused on signals intelligence.
In robotics, domain expertise and the early commercial leadership is generally concentrated where there is preexisting domain expertise in electronics and advanced manufacturing—in countries like Japan, South Korea, and Germany. Yet even as the industries of the future offer new opportunities to rising hotbeds of innovation around the world, it’s interesting to watch how Silicon Valley’s influence lingers and continues to draw start-ups in almost every industry. Consider the example of digital currency and fintech, industries of the future that blend old world and new.
New York and London are the world’s two dominant centers of domain expertise in global banking today, and both are home to substantial fintech investment. Over the past five years, the United Kingdom and Ireland were home to 52 percent of all the fintech financing in Europe. And New York drew even larger levels of fintech financing than London, with dozens of deals putting hundreds of millions of dollars into the bank accounts of technology companies trying to make the banking sector smarter.
But when Zac Townsend wanted to start a company focused on smartening up the banking industry, he did not start it in London or New York. He started it in California. It mattered less to Zac that expertise in banking specifically was in New York or London than that the expertise in innovation and its supporting culture was in California. He believed that in order to change the banking system, he had to work with it but away from it—an approach that is also reflected in the broader data.
Silicon Valley: the next Roman empire?
When twentysomethings like Zac decide to start a company and determine that in order to do so they need to be in California, it creates a self-perpetuating cycle. More broadly, Zac’s decision to base his new data-driven finance company in Silicon Valley reflects a roiling debate over just how domain expertise will develop in the big data industry and what effect this is going to have on the global economy as a whole. With the major impact that big data is having on almost every industry, the way that big data expertise develops has the potential to change the very nature of business. And investors are placing big bets on two very different answers.
Will big data serve to centralize businesses, pulling more industries into the gravitational field of Silicon Valley? Or will it allow more businesses to innovate wherever they are, in effect creating more opportunities in more places around the world than has been possible before now?
On one side of the argument is Charlie Songhurst, who sees the Valley as a burgeoning global empire. Because of Silicon Valley, Charlie says, “Global regional inequality is going to be unlike anything we’ve ever seen except maybe the comparative power of Rome versus the rest of the ancient world.” While I think Charlie overstates things, he makes an argument that’s worth examining.
His thesis lines up with a number of other thinkers who believe that Silicon Valley’s expertise in software and analytics will swallow up entire industries and cause massive centralization. The founders of Uber had no particular expertise in transportation, but that did not matter because of their ability to build a software and analytics platform. The idea that underlies Songhurst’s vision is that Silicon Valley companies could eventually run everything in which software and big data are useful—which is basically every industry on the planet. So what’s going to happen in this new data-driven empire, according to Charlie? “It’s a very simple equation,” he says. “Countries with high education and low wages will export IQ. That will be the Baltics, India, China. Of course, it’s terrible if you’re in Ohio or England or France or someone competing with an Estonian. So what you’ll get is a massive mean reversion of income throughout the world where the Valley, Israel, China, and maybe a couple of other places get very high economic returns and everywhere else in the world starts to revert to the mean. Again, it starts to look more like the Roman Empire.”
It is also the case that while the powers that be in Silicon Valley might not be the earliest movers in fields like precision agriculture, once success is achieved elsewhere, they don’t just sit back as passive bystanders and watch it grow. Google chairman Eric Schmidt recruited an Israeli entrepreneur, Dror Berman, to move to Silicon Valley and head up Innovation Endeavors, a large venture firm that invests Schmidt’s money. Israel is home to many of the 20th century’s great innovations in farming. Berman brought the intellectual curiosity about agriculture with him to Silicon Valley and developed Farm2050, a partnership that aspires to combine data science and robotics to improve farming with a group of partners as diverse as Google, DuPont, and 3D Robotics.
Dror recognized that Silicon Valley can be a little too navel-gazing, and told me that 90 percent of the region’s entrepreneurs focus on 10 percent of the world’s problems. With Farm2050, he is trying to bring Silicon Valley’s A game to agriculture. Silicon Valley’s history as a home to apricot and plum orchards is long past, and if it does establish itself as the source of winning investment or innovation for precision agriculture, it will contradict the idea that domain expertise will drive the industries of the future. Instead, it would suggest, as futurist Jaron Lanier has argued in his book Who Owns the Future?, that those who hold the most data, the fastest servers, and the most processing power will drive all growth from here on out.
It’s basically the idea that Google could do my job and your job— and everyone else’s job—better if they wanted to simply by applying their top-of-the-line analytics abilities. There is an increasingly large audience, however, that holds a different view from Charlie Songhurst. They believe that big data, instead of absorbing and supplanting other industries, will serve as a broad tool that every existing industry can use to spur growth. The idea is that data will become widely usable and scalable enough that it won’t have domain expertise in the same way that other high-barrier-to-entry industries of the future like genomics or robotics do.
This view was explained to me by Mark Gorenberg, a veteran of West Coast venture investing who saw the investment case for analytics early and built a venture capital firm around it, Zetta Venture Partners. Mark has been in venture capital for a quarter-century and splits most of his time among investing, work with MIT, and serving as an advisor to the president of the United States as a member of the President’s Council of Advisors on Science and Technology. Gorenberg believes the big data economy will extend far beyond Silicon Valley.
He says: “Analytics businesses will come from anywhere. You have the algorithmic expertise on one side that is coming out of a lot of universities and you have the domain expertise for particular industries, which manifest themselves everywhere.” Gorenberg argues that as the big data market grows over the next couple of decades, it can be a source of revitalization for old industrial centers where local domain expertise exists.
In the rust belt, for example, he sees strong opportunities for the development of analytics firms rooted in the region’s strength managing industrial processes. He sees Boston’s strength as a biotech center enabling it to create health data companies, and he foresees the creation of energy analytics companies in Texas. He predicts that we will see strong privacy and forensics companies forming around Washington, D.C., that build on the capabilities in the law enforcement and intelligence communities, providing high-paying jobs for people who worked at the NSA, CIA, and FBI.
If Gorenberg is correct that “domain expertise is everywhere,” then there is no reason not to be optimistic about the prospects for big data firms developing outside the United States. It just takes a combination of algorithmic expertise and domain expertise.
For example, having lost out on the wealth creation built around the Internet, Germany is now determined to leverage its domain expertise in logistics and household appliances to own the analytics markets in those areas of traditional strength, an initiative it calls Industrie 4.0. If the big data market develops as Gorenberg predicts—with the best companies being headquartered all over the world—then wealth creation will look entirely different for big data than it did for the Internet, where the benefits were concentrated in a 30-by-15- mile area.
I was persuaded by Gorenberg’s view when I was in New Zealand and saw an example of how the combination of big data and domain expertise will determine the geography of the industries of the future. New Zealand is home to twice as many dairy cattle as human beings. The Kiwis know cows.
Although New York and London are the global centers for banking, they are respectively second and third in fintech financing behind Silicon Valley, which gets about one-third of all the investments that take place in the fintech space. This raises an interesting question about just how distributed the industries of the future are going to be.
While there, I learned about the impact of Pasture Meter, a precision-agriculture technology developed in Palmerston, a community of 82,000 people in the Manawatu-Wanganui region of New Zealand’s North Island, more than 10,000 kilometers away from Silicon Valley. Pasture Meter uses advanced sensor technology to take 200 measurements per second over vast swaths of farmland to identify how much grass is in a paddock so that dairy cows can be distributed most effectively for feeding. It alerts farmers to the amount of feed they have and identifies low-production areas that need intervention from the farmer, say, more fertilizer.
Traditional technology for evaluating fields, like ultrasound or rising plate meters, typically captures only 250 readings of a pasture, whereas Pasture Meter can take up to 18,500 readings. Anyone with a phone can use the technology, and it works regardless of factors like weather.
It may seem that monitoring pastures—literally watching the grass grow—is an unnecessary use of real-time analytics, but the Kiwi farmers know better. With the massive increase in upward economic mobility in China, there was increasing demand for beef and dairy products, but New Zealand’s cattle farmers needed higher levels of efficiency—greater scale and lower prices—if they were going to sell into such a large market.
China’s population is 288 times the size of New Zealand’s. With strong domain expertise in dairy farming, local farmers and farm-equipment manufacturers knew that if they could feed their cattle more efficiently, that would increase output enough for them to export to China. What happened? Sales of beef from New Zealand to China soared 478 percent in one year. China surpassed New Zealand’s neighbor Australia as New Zealand’s largest export market, more than twice the size of what is exported from New Zealand to the United States.
The fact that it surprised me that it could be so important to know the location and concentration of grass in a field is exactly the point: New Zealand’s farmers had the domain expertise, so they knew what needed to be built and they built it. It’s too much to say that Pasture Meter deserves credit for the full 478 percent increase in beef exports, but local farmers bring it up as one of the important factors.
What happened in New Zealand can and will happen in other industries where there may not be a deep history of big data and analytics, but where there is domain expertise in another industry that knows where and how analytics would add value. The big data applications themselves are easily scalable, can be done broadly around the world, and can be implemented whether or not there is much previous data experience—as was the case for the people in Palmerston, New Zealand, who made equipment for the region’s dairy farmers.
Silicon Valley builds things that Silicon Valley wants, from nicer taxi services to more photo-sharing apps. But investors and entrepreneurs in Silicon Valley don’t see the world through, say, the eyes of farmers. Thus, they are less likely than a company in the Manawatu Wanganui region of New Zealand’s North Island to recognize the need for and develop a technology that enables greater beef and dairy production for export to China. While Marc Andreessen is as closely identified with Silicon Valley as anydiv, he agrees that those fields that are early in their development can and should take root wherever there is deep knowledge about a specific area.
He has proposed that Detroit leverage its expertise in automotive mechanics to become “Drone Valley,” and suggests that rather than trying to create more Silicon Valleys, we should hope and plan for the creation of “50 different variations of Silicon Valley, all unique from each other and all focusing on different domains.”
Larry Summers reinforced Andreessen’s view, telling me, “My general line is, in essence, there is much more division of labor than there used to be. Strategies for countries, companies, and people are much more about building on your strength than offsetting your weaknesses than it used to be.”
This effectively means to stop trying to chase after Silicon Valley and focus on the skills and processes that will unleash the next wave of innovation in fields in which there is already local expertise. My view is that the geographic spread of domain expertise in the industries of the future will ensure that the next stage of globalization produces centers of innovation and commercialization that are more geographically diverse than the last stage, when Silicon Valley enjoyed 20 years of domination.
There won’t be a Roman Empire. The thought that software and big data–savvy companies and entrepreneurs in Silicon Valley will reign supreme is not a crazy one, but I think that as big data becomes more widely adopted, it will be more of a commodity that any industry can use. There is a real opportunity for stakeholders with domain expertise to innovate for themselves, and in breathtaking ways.
But if they wait too long, some 28-year-olds in California are going to do it instead.
In the cases where an industry is too slow to adapt, then eventually more efficient, less expert start-ups (with big data expertise) like Uber will step in and take down companies with decades of domain expertise.
In the famous words of H. G. Wells, “Adapt or perish.”