The following is an adaptation of a talk I gave at the Jacques Delors Institute’s Digital Europe conference in Berlin November.
At X, the moonshot factory, we aim to solve huge problems in the world with the help of technology. Now, I bet you’re thinking, “Great, here’s another American telling us how to do moonshots. Those Americans have been obsessed with moonshots since the 1960s…”
Except that I am not American.
I grew up in Berlin, went to university in Oxford, married an Englishman and lived in London much of my life. So I am most certainly European. I first went to the US as an exchange student in 1988 where I spent the entire year explaining that Germany was divided and why the Berlin Wall would never come down. (My only excuse is that this was before the internet, and local news did not talk about Perestroika). Luckily, I came back to Berlin in the summer of 89 otherwise I would have missed the most important event in my hometown’s history!
I have worked on and off for American tech companies since 1999, including the last four years at X in California. As I was preparing for this trip to Europe, I started thinking about the lessons I could share from my time in the Valley. Countless blog posts and entire books have been written about this topic. In fact, Silicon Valley culture has become such a cliche that there is even a TV series about it. So I started wondering if anyone had looked at what lessons the Valley can take from Europe.
Curious, I typed ‘lessons from Europe’ into Google. I got mostly economics articles, hardly anything about business. One of top results was entitled ‘Lessons from Europe for American Business’ published in 1963. “Right,” I thought, “time for an update”.
Today, I want to share some of the lessons I’ve taken from my work at X. But I also want to offer the reverse: lessons that Silicon Valley can and should learn from us Europeans. Now, more than ever, we need greater collaboration between the two sides. To solve the world’s largest problems we need the brightest, most diverse minds from all fields to work on creative solutions together.
I want to start by convincing you of something that may sound ridiculous: I want to convince you that it’s often easier to make something 10 times better than 10% better. If I say, “make a car that can drive 100km on 1 litre of petrol,” we might be able to get there eventually by improving current methods bit by bit. But if I ask you to go 1,000km on 1 litre of petrol, you have to totally reimagine things. You have to start over. This is not only very liberating but also more inspiring because other people will want to help you solve the impossible.
This is a belief that we’ve built into how we define the moonshots that we work on at X. We start with a huge problem in the world which affects millions or even billions of people. Then we propose a radical solution, a product that sounds like science fiction today. Then we work on inventing breakthrough technology which makes it possible to turn this from fiction to reality in the next 5–10 years.
Our first project was the self-driving car. Around the world, 1.2 million people die on the road each year and 94% of accidents are due to human error. The founder of the project was Sebastian Thrun, a German roboticist and Stanford professor who lost his best friend in a car accident when he was 18. Sebastian’s radical solution to the problem of road safety: a car that drives itself. A car that is never tired or distracted, never texting and never drunk.
In the 20th century, as you can see in this wonderful 1957 advertisement, self-driving cars were science fiction.
Advertisement from 1957 for “America’s Independent Electric Light and Power Companies”
I love that the family is playing dominoes. Today the parents would be checking their email and the kids would be on YouTube or Snapchat.
Now, we’re on the verge of turning that science fiction into reality. In the past 10 years sensors, computer vision, image processing and mapping have come so far that cars have shifted from being purely mechanical to being driven by software.
We have been working on self-driving cars at X for over seven years. This lengthy commitment is unusual in Silicon Valley, which prides itself on speed of execution. Get to beta as fast as possible, launch your app, get millions of users, IPO. In the frantic pace of the Valley, it can be hard to do long-term projects. However moonshots of this scale and nature require a long-term approach. This is something Europe understands.
In Europe, long-term thinking is taken for granted. My alma mater, Oxford University, was founded 800 years ago, Heidelberg University over 600 years ago, Stanford University just over 100 years ago. Europeans build their houses out of bricks, not wood, and expect to live in them a long time and pass them on to future generations. Europeans move 4 times in their lifetime, Americans nearly 12 times. Europeans embraced renewable energy decades before it became fashionable for Californians to put solar cells on their rooftop. European cities have learnt through hundreds of years of trial and error, build and rebuild, how millions of people can co-exist in a small space. German Mittelstand companies (small to medium businesses) are passed on through generations and their strategy is set in decades, not quarters.
Why does all this matter? Long term thinking means that investment decisions can be long term too. By taking the long view you give engineers and innovators the time and space to build a product that is truly transformative. The automotive, energy and transportation sectors all require long term investment plans. This kind long-term thinking Californians can be reluctant to commit to. Of course they would tell you that you can’t plan innovation…
The reality is that innovation is incredibly messy. Anyone who tells you otherwise probably isn’t innovating. If we knew the answers to the hard problems, they’d be solved already! The only way to find out is to launch early and iterate, rather than wait until something is perfect and never launch. This approach is easy with software, which you can constantly update. It’s not so easy with hardware, as we learnt with our self-driving cars.
Four years ago, some Google colleagues volunteered to help us test our self-driving cars. At the time, our cars could only handle autonomous highway driving. So our testers had to drive manually to and from the highway before the car could be put into autonomous mode.
This experiment taught us something critical: it was a fallacy to assume that drivers could take over if there was a situation that the car couldn’t handle. Even though our testers knew they were testing early stage technology and had promised to be ready to take over at any time (and were on camera!), they didn’t pay attention like they should have. One person even turned around looking for a phone charger on the back seat, while the car was going 80km per hour.
This insight led us to focus on full autonomy and creating a car that can drive itself from A to B at the push of a button, instead of just adding more and more driver-assist features. We began building prototypes without a steering wheel or pedals.
Self-Driving car prototype
There are valuable lessons to taken from the messiness, but it can be counter-cultural for Europeans who like to plan ahead. At X we often host city leaders from across the world who want to learn from us about self-driving cars. European cities want to know how to plan for the future and how the physical environment of roads and traffic lights, regulations and laws, should be adjusted so that new types of transportation can be accommodated.
It’s a bit of a cliche that Silicon Valley is better at embracing failure than Europe. Let me be clear: nobody aims for failure. All entrepreneurs in the Valley aim for success. If the idea of ‘failing fast’ feels counter-cultural, try re-framing it as ‘learning fast’.
In the early days of a project, I’m very focused on helping my teams figure out what’s the most important thing to work on first. We make a list of all the technical, product and business risks and we rank that list based on how serious they are. Would they sink the project? How confident we are that we can solve these problems? So far this isn’t anything special. Where we are different is that we run at the hardest problems first. It is very tempting to start by working on the problems that you know you can solve, because it gives you a sense that you are making progress. But if you’ve spent a year working on all the easy things, only to discover in 2 years time that there is a problem you can’t solve which sinks the entire project, you’ve wasted all that time and effort. We’d rather find that out sooner than later.
I ask my teams to make an honest assessment of all the reasons why it might NOT work. We also agree on a list of economic and technical “kill criteria” (things that make it legitimate to kill the project) upfront. This means that it’s much less emotional when we do decide we have to shut down a project.
Turning seawater into fuel with Project Foghorn
One of my favourite projects was Foghorn. The goal of Foghorn was to make carbon-neutral fuel. The team actually figured out how to solve all the technical challenges (at least theoretically), and we built a working prototype in partnership with Xerox PARC. We extracted CO2 from seawater, made hydrogen using clean electricity, and then combined them to make a liquid fuel. When it burns, the CO2 gets released back into the atmosphere, creating a closed carbon cycle.
The bigger challenge turned out to be economic: to make fuel at a price that was competitive with conventional fuel so it would be widely adopted. When we started the project, oil was at $100/barrel. When it went down to $30 the team decided to shut down the project. You can read more about their journey in Fast Company here.
I like to think that Project Foghorn isn’t dead, just deep-frozen. We have submitted a paper to a peer-reviewed journal. Hopefully somebody will read our paper and have an idea that we overlooked. Or maybe one day carbon emissions will be priced into the cost of fuel.
The challenge that Foghorn tried to solve is just one example of the inherently physical nature of some of the world’s biggest problems. These kinds of problems require physical solutions.
Silicon Valley started with silicon design and manufacturing, but its strength in the past few decades has been software. The Valley’s developers are among the most creative and prolific in the world. They also make it easy for others to build more software. In 1999 when I was at eToys.com, we hand-coded our ecommerce site and built our own warehouse and customer service software. We had to be on call over Christmas in case our servers fell over due to the spike in traffic. Now you can get all this as a software service and you can be up and running in the cloud in just a few days.
But many of the world’s really pressing problems are physical in nature and require physical solutions.
Clean water, renewable energy, transportation, healthcare, agriculture and connectivity won’t be solved with software alone.
Europe has centuries of experience in dealing with the physical world, and a proud history of engineering at universities and companies. With the rise of robotics and automation, Europe’s strength in electrical and mechanical engineering is now more relevant than ever. Software is what makes the machines intelligent, but they need sensors to understand the world around them and actuators to interact with it.
At X, we not only love to hire those great European engineers, we also regularly turn to European partners for their expertise. Which brings me to my next point.
The myth of the lonely genius inventor is just that: a myth. Steve Jobs didn’t make the iPhone, Elon Musk didn’t build the Tesla car, and Larry Page and Sergey Brin didn’t make the Google search engine all by themselves. Great innovations are made by great teams.
Silicon Valley startups have flat hierarchies because ideas can come from everywhere. At X, the team that works on the early stage ideas is called Rapid Evaluation. Their job is to generate lots of ideas while at the same time trying to kill these ideas as fast as possible, so we can focus on those that have the most potential. This team votes in ideas democratically and nobody from the X leadership level gets involved until that group has decided the technology they are working on is ready to move to Foundry, the next stage, where we ramp up investment in these new technologies.
As well as a flat structure, our teams at X have a mix of backgrounds. We have ex-Navy Seals working alongside fashion designers, firefighters and concert pianists. We bring together people from different cognitive backgrounds because they bring different points of view, which causes constructive friction. This kind of of friction helped us make a big change to Project Wing, our automated aerial delivery system.
Defibrillator user experience studies for Project Wing
Our first application for Project Wing was to deliver defibrillators to heart attack victims. However, when our user research team gave a defibrillator to bystanders, they found that it took people about 6 minutes to figure out how to use the defibrillator. After 7–8 minutes the ambulance would arrive anyway, making it irrelevant that we could deliver the the defibrillator in 90 seconds.
These researchers approached the problem in a completely different way: the engineers were focused on how can we make it work, the researchers were focused on why. Based on their insights, the whole course of the project changed.
What Silicon Valley does better than any other place is to define your team as broader than your company or immediate circle. When I first came to the Valley, I was surprised by the readiness with which people freely share ideas, offer up help and consider collaborating — even with people who look like competitors.
X relies on partnerships to make our products a reality. We’re good at core technology breakthroughs and creating working prototypes. However we recognize that we don’t have expertise to take products to market in a range of industries, so we partner with people that do. For example we have worked with Novartis on our smart contact lens, with Telefonica on Project Loon and with Fiat Chrysler on self-driving cars — these all happen to be European companies. If we can build the relationships and the trust, there is a huge opportunity for European companies to collaborate more with technology companies in Silicon Valley.
Now more than ever before, European cities like London and increasingly Berlin, are truly global melting pots. When my son was 3 years old attending nursery school in central London, his best friends were a girl with Scottish and Italian parents, an Indian boy, and a Somali refugee in a wheelchair. He was completely gender, colour and disability-blind because that was just his reality.
Silicon Valley has less diversity. There are not enough women or black people in tech and most people are relatively well-off. As a result, while many Silicon Valley founders do think big, there are also many solving problems that aren’t real problems. My heart sinks every time I see a poster in the Bay Area advertising an app for a laundry service, house cleaning service, or home chef service. It feels like boys building products for the things their mum used to do for them… My heart sinks even more when I see European startups copying these. Thankfully, that seems to be happening less these days.
In a world of global networks of technologists living in cities across the world, startups in Europe have the benefit of being close to the real needs, the real problems, and the real interests of consumers. If you live in a messy, old city like London, Paris or Stockholm, you’re cheek by jowl with people working in different sectors, from different backgrounds and with different needs. You share your lives, your schools, your hospitals and your transport with the rich tapestry of life.
In Silicon Valley — far away from most big cities, comfortable in your car and your private health care — it’s much harder to truly understand the needs of other communities. It doesn’t end with consumer products either. European universities are full of innovations in science and technology that are not being turned into startups and could and should be. (If you know of any that might be a fit for X, please let me know. Some great X projects had their origin in university research.)
While Europe and Silicon Valley are very different, there are lessons to be learnt from both cultures, and both sides would benefit from more collaboration.
Now let’s talk about why this matters: to solve the world’s largest problems, we have to bring together the brightest, most diverse minds to work on creative solutions.
People sometimes ask me, what technology are you most excited about? I could talk about advances in sensors, the cloud, machine learning, artificial intelligence and robotics. But these are just tools.
A better question to ask me is ‘what problem are you most excited about?’ My current answer is: climate change. We have to find ways to reduce the harmful impact humans are having on our environment. As a tech optimist, I do believe that technologies can contribute breakthroughs to manage climate change, but we can’t do this alone. The world needs an ecosystem of universities, labs, companies, investors and regulators working collectively to bring these technologies into being.
To provide clean energy, reduce carbon emissions, and remove carbon from the atmosphere we need radical innovation from all of us: scientists and engineers, entrepreneurs and corporates, investors, policymakers and governments. Together we can solve the biggest problems facing our generation, but we can only do it if the brightest minds from Europe, Silicon Valley and across the world collaborate. I hope you will join us on this quest.
If you have ideas or are interested in collaborating with X, please get in touch.