‘If we want to build truly intelligent machines, we will have to embrace humankind’
Will we ever be able to build a machine that is as smart as you and me? A ‘human machine’ that is conscious of its own existence? This is one of the topics that Marcel van Gerven, professor of Artificial Cognitive Systems at Radboud University, will touch upon during his inaugural lecture on 14 September.
He was initially a cognitive scientist and now develops variants of artificial intelligence (AI) that are inspired by biology. Additionally, he uses AI techniques to better understand the brain. Three questions for Marcel van Gerven on the occasion of his inaugural lecture.
Why do you observe the brain to improve AI?
“I want to understand intelligence in its form as it occurs in nature. I think that in the development of new smart algorithms, we should combine insights from social sciences and natural sciences to produce better AI. This could also help us gain new insights into the way the brain works, and at the same time realise an improved collaboration between man and machine, such as Brain Computer Interfaces.”
So, do you think we’ll ever be able to build human-like machines?
“Today’s AI systems still lack important qualities that are present in the brain, and many of the current breakthroughs in AI could be classified as “statistics 2.0”. I expect the AI forms inspired by biology to bring us closer to machines that can adapt to complex physical, biological and cultural processes in the world around us. I think there are no fundamental limitations that would impede us from ever building machines that equal the natural intelligence of biological organisms.”
How does an AI scientist deal with the worrying scenarios that are often outlined nowadays, such as the one with robots getting out of control?
“We are nowhere near getting there. Although it is certainly important to have research carried out into the long-term effects of developments in AI, I think it is currently more urgent to focus on the impact of AI on today’s society, like the privacy issues. Besides, I personally want to contribute to the positive applications of AI that are feasible already, like using AI to help visually challenged people to see again.”
“We should also not forget to keep working on fundamental research in AI: building more powerful and more efficient intelligent algorithms. The university should remain a refuge for curiosity-driven research. That is what we require to reach true breakthroughs; this also applies to AI.”
Marcel van Gerven will give his inaugural speech on Friday 14 September in the Aula of Radboud University. More information.
More information about AI research at Radboud University is available at www.ru.nl/ai