This research contributes to the following Sustainable Development Goals, among others:

Artificial Intelligence

Machines that learn, evolve and interact with humans

Automation started with the first computers, 70 years ago. Now we are entering the era of 'automation of automation': computers that learn from experience and improve their own software. The world is changing. Researchers from various disciplines at Leiden University are working on artificial intelligence that will enrich human intelligence.

Artificial Intelligence all around us

‘AI is nothing new,’ says Aske Plaat, Scientific Director of the Leiden Institute of Advanced Computer Sciences (LIACS). Together with other researchers at Leiden University, he’s committed to the development and application of Artificial Intelligence (AI). ‘Intelligence is all around us nowadays, even an automatic sliding door has some intelligence.’ 

Holger Hoos, Professor of Machine Learning at LIACS, thinks AI will become pervasive: ‘The way we do science now is driven by mathematics. In the future, it will be driven by AI.’  

In the last five years, there has been a revolution in AI, especially for neural networks: self-learning computer programs that mimic a network of brain cells. Thanks to better algorithms and increased computing power, the prospects for the next decade look even more spectacular. The networks will then be ‘fed’ with huge amounts of training data for these networks, such as images, video, audio, text on the internet.  

Bug-free software
Automated learning and automated reasoning are the mainstays of AI. Checking that software does exactly what it is supposed to do can now be automated, thanks in part to research in Leiden. Major parts of computer chips and an increasing amount of crucial software – in airplanes, for instance – is now rigorously checked by AI programs. Hoos: ‘We can guarantee that this software is bug free.’ 

Plaat and Hoos both expect AI to be of benefit to personalised health. Recognising patterns in large amounts of patient data could lead to therapies that are tailored to individual patients. AI will also help design molecules that are better medicines, with fewer side effects. Leiden’s computer scientists are already working with the pharmacology department of the LUMC, the University’s teaching hospital.

Plaat: ‘In Leiden, we are also strong in computer vision. Besides medical applications, this will benefit a wide range of areas, from archaeology to self-driving cars.’ The Faculty of Archaeology is using AI for archaeological research and in the hunt for burial mounds in the Netherlands. Researchers at the Faculty of Law are looking at the implications of AI for society and ethics: what happens if algorithms start making choices for us? 

Leiden is also strong in computational linguistics, natural language processing and speech recognition. LIACS and the Faculty of Social Sciences and the Faculty of Humanities work together closely in this area. It is the presence of and cooperation with so many disciplines and fields of research that make the AI research in Leiden so interesting. This also attracts researchers from elsewhere.

Keep up with the US and China
Holger Hoos is a co-founder of CLAIRE, a European network of AI laboratories. Their mission: to get Europe to invest in AI on the same scale as can already be seen in the US and China. Hoos: ‘Do we want American and Chinese companies to be selling us the AI technology of the future? We need to reach critical mass in our AI research, and for Europe to act as one on this.’

Leiden Institute of Advanced Computer Science
Faculty of Science
Faculty of Social and Behavioural Sciences
Faculty of Archaeology
Faculty of Law