The Future of Dutch AI: An Interview with Professor Cees Snoek

Tuesday, 24 September 2024 at 19:45

Professor Cees Snoek, professor of AI at the University of Amsterdam and scientific director of Amsterdam AI, shares his take on where artificial intelligence stands today—and where it’s heading. We spoke with him about breakthroughs, hurdles, and how academia and industry are teaming up.

He’s enthusiastic and driven, but he also sees trouble for the Netherlands: “We need AI-ready supercomputers, and that takes roughly half a billion euros.”

Can you tell us about your background and how you got into AI?

“My interest in AI started during my computer science studies. I was fascinated by how we can teach computers to handle complex tasks, like recognizing objects in images or interpreting natural language. That drew me into computer vision and deep learning. I now lead a research group at the University of Amsterdam focused on AI for video and image recognition.”

What do you see as the biggest AI breakthroughs in recent years?

“Without question, the rise of deep learning has been a game changer. It enabled us to process massive datasets and learn patterns that were previously out of reach. Think self-driving cars, advanced medical diagnostics, and even AI that can create art.”

And what breakthroughs are you seeing right now?

“Sora is fascinating. OpenAI has surprised us again by showing they can move far faster than the rest of the world. It underscores the pace and power of AI’s evolution. The speed of progress is striking.”

What challenges do you see for the future of AI?

“A key challenge is how we interpret and explain AI decisions. We need systems that are transparent and understandable, especially in critical areas like healthcare and justice. We also have to take ethics and societal impact seriously—privacy, bias, and accountability.”

How will AI shape our daily lives over the next decade?

“AI will be woven much deeper into everyday life. It won’t just assist us—it will make decisions autonomously across a wide range of applications. From personal assistants planning our days to intelligent systems optimizing energy use and nudging us toward healthier choices.”

What does “Responsible AI” mean to you?

“Responsible AI means accounting for real-world impact. It’s about ensuring systems aren’t biased, don’t exclude people, and don’t overuse compute resources. We must deploy AI responsibly—so it benefits society.”

How do you collaborate with industry on AI?

“We run multiple research labs in Amsterdam together with companies. I’m involved in labs with Qualcomm for mobile, TomTom for high-definition navigation maps, and Core42 for medical imaging. We conduct joint research, and they fund a large share of AI research in Amsterdam. That partnership helps us build new methods and technologies that push both science and industry forward.”

What are the infrastructure challenges for AI in the Netherlands?

“The Netherlands must invest in its own AI infrastructure. That requires significant resources, and government leadership is crucial. Without it, we risk falling behind the global AI race. We need AI-capable supercomputers—an investment of about half a billion euros. Collaboration with European partners can help.”

What role can companies like ASML play in building Dutch AI infrastructure?

“Companies like ASML can help fund efforts and prioritize delivery of GPU chips. But the primary responsibility lies with the government to invest in a national AI infrastructure. It’s vital that this infrastructure stays in public hands to ensure independence and broad access.”

What worries you about the future of AI research in the Netherlands?

“I’m concerned we’re falling academically behind industry. We don’t understand large-scale models well enough and can’t reproduce them due to limited compute and resources. That means we can’t train students on the most advanced technologies and risk slipping in innovation.”

What must the Netherlands do to secure or claim leadership in AI?

“Keep investing in AI research and infrastructure. Work closely with European partners and build a national AI strategy. Continuous innovation and long-term investment are essential to stay ahead in AI’s fast-moving landscape.”