‘The Netherlands falling behind in artificial intelligence’

Self-driving cars, smartphones, robots in healthcare: artificial intelligence (AI) is everywhere. And the developments in the field are continuing at breakneck speed. If the Netherlands wants to continue being relevant in the world of AI, it will need the requisite expertise at home. But here lies the problem.

In one year’s time, the Netherlands has slipped from fifth to fourteenth place in Oxford University’s AI Readiness Index, which ranks national governments in terms of developing the potential of AI. The Dutch Research Council (NWO), an important funder of science and research, acknowledges there is worldwide competition for AI talent. ‘The Netherlands faces a choice: do we wait for AI to take shape without us, or do we want to be part of that shaping process?’

Five past midnight
NWO believes in the second option and is drawing up a national artificial intelligence research agenda, which is supported by the entire academic field (the arts, sciences and social sciences). The agenda describes all aspects of the development of an AI algorithm or AI system. And not just the technical aspects, but also the interaction of AI with people and society.  

“It’s not five before midnight, but five past,” says professor Inald Lagendijk, AI expert attached to Delft University of Technology and chair of the authoring committee of the research agenda. He says it is high time to take action, TU Delft has extensive plans in the making as well, even if that action is somewhat belated.

Given its small size, the Netherlands is still in a reasonable position. What are we good at? “We’re good at machine learning, which is the development of algorithms and technologies that allow computers to learn. But we’re also good at teaching computers ethics. Take self-driving cars, for example. You can teach them the traffic rules, but there are also moral values that are trickier for systems to master.”

Why is the Netherlands falling behind?
“It isn’t the case that the quality of research into AI is falling; in fact, it’s still pretty high. But other countries are speeding up the pace of their research and are prepared to invest more. There is a real race for talent going on. China and the US, but also countries closer to home such as Germany and Sweden, offer far better employment and working conditions.”

Starting package
“Entry-level scientists earn more at large American corporations than professors. And they also get money to purchase a home in the city. Germany has made a starting package of €5 million available for each of its thirty new AI professorships. It’s utterly inconceivable that Dutch universities could offer packages like that. So it will be difficult to attract and retain good talent.”

One country can’t be expected to excel at everything. Is it really that bad if the Netherlands loses ground in this area? “The impact of AI on society, politics, the economy and science cannot be overstated; that’s what we need to focus on. Falling productivity means we lose expertise. If that happens, we will no longer be able to train future generations because we will not have enough scientists. We can already see this happening in the current enrolment restrictions for AI and Informatics degree programmes. Huge numbers of students are keen to take these degrees and there is plenty of demand from the market, but we simply cannot keep up.”

People-focus
“Moreover, failing to invest in AI at home means we will no longer possess the expertise to work with AI responsibly. We will be at the mercy of the whims of foreign players, such as China – where government involvement is ubiquitous – and the US – where commercial corporations like Facebook and Google call the shots. Europe wants to go a different route, one where technology is more people focused and the ethical side of AI is taken seriously.”

We are already being overtaken by events?
“Look at Facebook and similar platforms that are fully American owned. Our influence on them is so darned insignificant, while they are impacting our society. These concerns are also shared on other areas, as evidenced in the debate on the security of our networks. In that field, we are mostly dependent on Chinese and Russian services such as Huawei and Kaspersky anti-virus software.”

What is the value of this agenda?
“The aim is to counter fragmentation of AI research and guide the efforts of government and the business sector: in which areas can they join forces and what is going to be the Netherlands’ focus in AI? But this is all still in the discussion stage. Eventually, we will need to exchange words for deeds. A number of initiatives have already been launched, such as the Strategic Action Plan for AI and the Dutch AI coalition, which brings together the government, knowledge institutions, and trade and industry.”

What action would you like from the government in The Hague?
“We need €150 million in private investment and a further €150 million in public investment each year if we are to equal the efforts of the surrounding countries. But I can’t say if that investment will actually come. Intensive discussions have already been held with members of parliament, Secretary of State for Economic Affairs Keijzer and Prime Minister Rutte, and these discussions are expected to continue in the months ahead. But it’s far from a foregone conclusion; we still have no guarantee that investments will be made.”

HOP, Melanie Zierse
Translation: Taalcentrum-VU