SAB member Professor Nicu Sebe: “Human being should be the focal point in hybrid intelligence”

Professor Nicu Sebe shared his thoughts on the most interesting aspects about hybrid intelligence when he visited University of Oulu. Sebe is a Professor of Computer Vision, Multimedia and Affective Computing at the University of Trento, Italy, and a member of international Scientific Advisory Board (SAB) of Hybrid Intelligence programme at the University of Oulu.
Professor Nicu Sebe at the University of Oulu gardens

Nicu Sebe is a Professor of Computer Vision, Multimedia and Affective Computing at the University of Trento, Italy, and a member of international Scientific Advisory Board (SAB) of Hybrid Intelligence programme. When he visited the University of Oulu, we interviewd him about his definition of hybrid intelligence.

"Hybrid intelligence is equivalent to human-enhanced, or human-supported, intelligence”

When asked his definition for hybrid intelligence, Professor Nicu Sebe sees the term equivalent to human-enhanced, or human-supported, intelligence.

“What I want to say with that definition is that the human being is and should be the focal point. What we do with AI is to enhance human abilities and to provide services for the benefit of people”, says Sebe.

It doesn't necessarily mean that a human needs always to be in the loop, but it means that if AI takes independent decisions, they need to be taken considering the humans.

“Explainability and interaction are the most important aspects of hybrid intelligence at this moment”

There are two aspects that Sebe finds especially interesting in the research of hybrid intelligence. These are explainability and interaction.

“By explainability I mean that it is extremely important to be able to understand a particular output or decision of the AI system. I find it less critical if AI makes mistakes as long as we can understand why these are made, e.g., the case of biases”, says Sebe.

Sebe thinks we should go beyond human-readable explainability and should demand of the AI system to realize what it is doing and to be able to learn from errors or from human examples.

Interaction is related to the first aspect in the sense that we will want to have a seamless interaction between the AI system and the human, explains Sebe. In this respect the AI is regarded as a companion which not only listens and learns from the user but can also generalize and intervene when needed.

“I see this AI to human interaction to be very close to what human to human interaction is expected to be”, Sebe concludes.

“The Oulu HI team has a nice subject and good potential - I hope the ambitious goals will be successfully achieved”

Visiting the HI programme at the University of Oulu for the first time at the end of May 2024 was very interesting to Professor Sebe, since he has been thinking about these subjects in his research for more than 15 years now. One example is a white paper called Human-Centered Computing Towards a Human Revolution, written already as early as back in 2007.

“I'm a technical guy but was always fascinated by the human-centric approach”, tells Sebe.

When asked about his expectations for Hybrid Intelligence research programme, Sebe sees that researchers at the University of Oulu have taken on a very challenging endeavor, since interdisciplinarity is never an easy task. The main difficulty is putting together all the disciplines and making the researcher work in a truly interdisciplinary environment.

“This doesn't mean that each one of them should not excel in its own field - but rather the fact that one activity done by a group is conducted with the requirements of the other groups in mind.”

In this respect, Sebe wishes the programme would be able to achieve important steps in this direction. Sebe thinks the team has a nice subject and good potential and he hopes the ambitious goals will be achieved.

Last updated: 9.8.2024