Artificial intelligence - An opportunity for Europe
Around the turn of the millennium I met Kenny Hirschhorn in London. I was working as a consultant on the merger between Orange and France Télécom. Kenny was a futurologist at Orange; they didn't have a strategy department back then. He held a cell phone up in front of me and said: "What do you see?" "A cell phone," I replied. Wrong answer. After a few more failed attempts at guessing, he enlightened me: "This is going to be the remote control of your life." He played a video from the United States showing a businessman unlocking his car with his cell phone, coordinating his family's appointments and receiving e-mails on the go. I was impressed. That was in January 2000.
In the summer of 2017, more than 100 entrepreneurs and researchers from all over the world wrote an open letter to the United Nations warning of the threat from artificial intelligence. AI, they said, may be repurposed to develop deadly "autonomous weapons" and could become the "third revolution of warfare". Elon Musk and Mark Zuckerberg have even argued vigorously over AI recently. The founder of Tesla and Space-X accused the Facebook boss of having a "limited understanding of the new technology". Zuckerberg branded Musk a naysayer, claiming he was fanning the fears. So who is right? And what would Kenny have to say about "killer robots"?
The problem is not that Musk and Zuckerberg disagree, but that they are not talking about the same thing. Both of them are referring to artificial intelligence but they're thinking in such different timeframes that their statements are simply not comparable. When Zuckerberg speaks of the positive potential of AI, he has mainly the next ten years in mind. Even Kenny would have been overwhelmed by today's possibilities. In 2013, there was about as much data in the world as known stars in the universe. By 2020, the global volume of data will be ten times greater.
From my own perspective as a global management consultant, AI offers enormous opportunities for businesses and for society. Like Stanford professor Andrew Ng, who until recently led the team of 1,300 AI researchers at Chinese search engine company Baidu, I think we're facing similar upheavals to what the first technical applications of electricity brought a century ago. AI uses algorithms to generate logical insights and make forecasts. Depending on how it is programmed, it can be used for critical decision making. This will have a groundbreaking impact. For many people, AI represents the hope of a better life. Self-learning machines will help doctors diagnose illnesses and select the treatment within seconds. The AI start-up Sentient AI, for example, is able to detect blood poisoning with a 91 percent probability – 30 minutes before the first symptoms are felt.
Kenny would also be thrilled by the innovative actions of the many companies working at different locations around the globe to develop AI solutions. New centers of gravity are emerging, the rules of the game are changing. As far as I'm concerned, it is a positive thing that the leading AI experts are no longer working for Google and the like but are doing independent research and want to set up their own companies. In California, young developers are already moving from San Francisco to Los Angeles and New York. This is where new hubs are being created, fostering the decentralized development of the technology.
For Europe, too, making a name for ourselves as an alternative to Silicon Valley is a huge opportunity. But what we need is a research and start-up initiative to accelerate the creation of European centers of artificial intelligence. Because part of the wider context is about defining the global standard for this technology. That is an area where Europe used to be proactive, having been instrumental in the establishment of the GSM standard, for example, through the European Telecommunications Standards Institute.
Now is the time to take the next bold step. If Orange had listened to Kenny, the smartphone might have been invented in this country. And that is why all of the researchers, entrepreneurs and policymakers involved here in this domain should act in unison today. With an Open AI initiative, Europe could become a leader in the development of portable and programmable AI devices. Because just as we carry a handheld computer called the iPhone around with us today, which possesses as much computing capacity as the entire US military had in 1984, within the next ten years we will be using completely new kinds of AI-based technology to help us in our everyday lives.
My personal vision is that the portable AI devices will run over private clouds – giving us back more control over the "remote control of our life". Protecting our identity and our data on platforms will be less problematic as we'll no longer rely exclusively on the offerings of Google, Apple, Facebook or Amazon. Our personal AI assistants will learn everything about our preferences, they'll anticipate and solve our everyday problems autonomously. They'll make purchases, book tickets, do our admin. We'll no longer be actively managing these tasks, but merely setting the targets. The rest will run in the background – at the best time and the best price. And this is precisely what we should be setting EU standards for.
I believe it is important for us to be promoting AI applications for the benefit of society at large. I'm thinking about education and cybersecurity. Public funding should target the artificial intelligence avant-garde and encourage the establishment of a transnational AI economy. Our start-ups need financial support, but they also need to be operating within a secure framework and, above all, they must have a free hand in how they act.
Microsoft and Facebook are showing us the way: They are standardizing their AI frameworks. This is how standards are set. Because AI has a greater significance than the mere technology behind it: artificial intelligence is the Industrial Revolution of the 21st century, heralding the wholesale transformation of the economy and society – a change that will be more radical than anything we have seen before.
The history of AI began in the United States in the mid-1950s. The first higher forms of AI application are not expected for 20 years. That is why we should tackle what will be happening in 2020 before we devote ourselves to fearing what's to come in 2038. And only if we act now to address the development of AI and the possibilities it brings can we also mitigate the risks mentioned by Elon Musk at an early stage. Let's take Kenny's example and throw ourselves into AI with all the innovative power we can muster. And let's look at it as a positive thing, because AI can bring us benefits in the near future that we cannot even imagine today.