“Universities must get ready for new types of teaching”

From traditional teaching to lifelong learning – everyone knows that we need to make the shift. But how? Pontus Wärnestål often grapples with that question.

Three people sits around, integrates with each other, with a table with computers and papers on it.

“People change career and we must be ready to support that. We need to develop new pedagogical approaches, platforms and methods.”

Pontus Wärnestål, Associate Professor in Informatics at Halmstad University

“All universities are designed for young people spending four years in academia, then going out into the world. But this doesn’t match the needs of society anymore. We must find the best ways to continue educating people parallel to their work”, says Pontus Wärnestål.

Pontus Wärnestål is an Associate Professor in Informatics, researching people-centered AI and digital service innovation. He is a design director at the company inUse and author of the book Design of AI-powered services. At Halmstad University he works with the project AI.m and the programmes MAISTR and Digital design and innovation, and he was a project manager at the founding of DLC, the University’s Digital Laboratory Center, where digital technologies for teaching and research are developed and tested.

In 2019, Pontus Wärnestål created a university course in the shape of twelve podcast episodes, Human-Centered Machine Learning.

“Lots of people use podcasts to learn things while commuting or exercising. I realized that this is a great opportunity for lifelong learning. On podcast platforms the specialty programs are seldom quality-assured, and the podcasts issued by academia are almost all produced by the communication departments”, says Pontus Wärnestål.

A man looks into the camera and smiles.

Pontus Wärnestål is an Associate Professor in Informatics at the School of Information Technology at Halmstad University.

Pontus Wärnestål figured there was room for something new; a real university course. With financing from Vinnova he and his colleagues set about creating such a course.

“We did much better than I had dared to expect and got over 6 000 listeners! Compared to a regular MOOC (Massive Open Online Courses), that is a lot”, says Pontus Wärnestål.

He sees an untapped potential in podcasts and pre-recorded lectures.

“The classical monologue lecture in a classroom is a waste of resources, I’d say. It forces students to become passive recipients, instead of active learners. At basic level, every teacher says almost literally the same thing as fifty others. It would be better with basic online courses for everyone, then branching out on site”, says Pontus Wärnestål.

Automation only half the point

Pontus Wärnestål has always enjoyed teaching and has received the Excellent Teacher Pedagogical Merit at Halmstad University. He believes that many more people need to be educated on AI and design. A common misconception, he says, is that AI primarily is useful for automating existing workflows.

“Anyone who believes that, misses at least half the point. You risk trying to automatize people out of the equation. This simply passivates the employees, they lose skills and knowledge, and quality goes down. We must see AI as a tool to strengthen people’s abilities, not a way to replace them”, says Pontus Wärnestål.

He exemplifies with an AI-enhanced camera developed by a tech company, which hired Wärnestål and some colleagues to analyze a possible implementation. The camera was meant to visualize basic health data in the emergency room, saving time for the nurses. Instead, the analysis showed that the basic idea was wrong. The simple fact that a nurse put a hand on the patient’s shoulder provided important information which the camera could not detect. It also comforted the patient. Rather, he medical staff and the camera needed to work together, measuring different things and increasing the nurses’ skills and competence.

Adjusting to the need for lifelong learning

There used to be a dystopic vision of super- intelligent AI taking over the world. This is not something that concerns Wärnestål. If anything, he believes “stupid” AI might pose a risk.

“Machine learning is always based on historical data, biased data, and machine learning reinforces bias. If we are to recruit a new manager and enter CVs from previous managers, the model will not tell us whether an applicant is a good manager. It will only say if an applicant is similar to previous managers. People are not very aware of this today”, says Pontus Wärnestål.

Pontus Wärnestål wants to continue exploring how higher education can adapt, to help with life-long learning.

“People change career and we must be ready to support that. We need to develop new pedagogical approaches, platforms and methods. Maybe completely re-shape our model – not reside only on campus but instead work part-time as researchers, part-time as learning coaches in workplaces. But today, we are still very conventional”, says Pontus Wärnestål.

Text: Lisa Kirsebom
Photo: Anna-Frida Agardson
Top illustration: Shutterstock


The Center for Applied Intelligent Systems Research (CAISR), is a research and education center for AI at Halmstad University. The center is funded by the University and the Knowledge Foundation with support from Swedish Industry.

The scientific focus for the Center for Applied Intelligent System Research (CAISR) is “aware” intelligent systems – human aware, situation aware and self-aware. Such systems can combine different sources of information to get an overall picture and monitor themselves. The subject expertise in the center is in signal analysis, machine learning and mechatronics. CAISR also has an emphasis on cooperating systems, in line with the research focus at the School of Information Technology (ITE).