Over the course of his distinguished career, Arnold Smeulders has made a lasting impact on the field of computer science by being pivotal to the development and large-scale application of image recognition software. The crowning achievement of his work came in 2014 when his spin-off company EUvision, which develops image analysis software, was acquired by global chipmaker Qualcomm industries. Published by UvA Persvoorlichting
Smeulders, professor of Multimedia Information Analysis, has been closely involved in the UvA’s Informatics Institute for over 25 years as a researcher, lecturer and former director.
What will be the biggest development in ICT in the coming years?
To predict the future, look to the past. Before 2000 the amount of available data was relatively small. This changed dramatically with the growth of sensors – from phones to eye scanners to cameras – which over the last five years has resulted in a flood of personal data. Such data, made up of digital numbers, can reveal key facts about an individual’s health, physical features or even socio-economic status and could be particularly useful in areas like healthcare. Once we begin to truly exploit these numbers and use deep learning to connect raw data and goals, many routine tasks, professional and personal, will be automated and enhanced. In future, the data will speak to us.
No, I don’t believe data will dominate us. Many fundamental features of human life have remained largely unchanged for most of written history. For example, a book like the Diary of Lady Murasaki, which portrays court life in imperial Japan in the 11th century, portrays a human existence almost identical to our own, one filled with all too recognisable emotions and relationships. The only real difference is technological. Which leads me to believe that although technology might change the way we do things, it won’t essentially change us.
It is true that decision-making is extremely crucial to law and healthcare. But it also equally true that such crucial decisions by their very nature require human agency. Would you want to be judged by something that removes all emotion from the equation and only takes account of the facts without grasping the context? Or in the case of healthcare, would you rather be told you’re dying by a mechanical voice or by a compassionate, flesh and bone person? Human-to-human interaction is essential to our existence. I can’t see that changing in anytime soon.
Once again, the past points to the future. Fifty years ago, a secretary spent most of her time typing. Today, this part of her work has been automated and has freed her up to focus on other crucial tasks like organising an office. In other words, she still forms an integral part of an organisation, but her role has changed. Or to use another example, when steam replaced the horse and carriage it didn’t only shorten the time spent travelling between places, but also vastly transformed the scope and purpose of travel and by extension opened up new areas of activity.
So, in that respect, I don’t take a cynical view of innovation. Technological change alters but also improves human existence.
The absence of enforced central steering. In over 25 years at this university, I cannot remember a single instance when someone told me how to do my research or teaching. This shows how much value this university places on autonomy and individual initiative. Of course, the UvA has strict requirements for those want to work or study here, but it also offers a lot of room to experiment and develop ideas. From personal experience, such responsibility and faith only bring out the best in people.