Thinking as a Superpower for Learning – Even in the Age of AI

The Marsio Presents event series is organized by LES Teacher Services. It brings together insightful presentations from across Aalto University and invites the audience to listen and take part in the discussion. The aim is to increase knowledge sharing, inspire the development of teaching and learning practices, and strengthen dialogue within the community.

Vice president of Education Petri Suomala opened the event by noting that the right moment to discuss the future of learning is now. The changes brought by AI offer an opportunity to rethink the future of teaching: if something can be learned independently or with the help of AI, what is worth teaching at the university?

University education has traditionally emphasized knowledge. In his remarks, Suomala asked whether other dimensions of learning—skills, attitudes, and habits—could be highlighted more strongly alongside it. For example, emphasizing motivation, autonomy, and active learning are good cornerstones as we consider the next steps in defining learning today and tomorrow.

Human memory and learning: implications for university teaching

In their keynote, professor Iiro Jääskeläinen advised that teachers should focus on clear goals and transitions, coherence, taking students’ prior knowledge into account, and encouraging students to ask questions. It’s important to revisit key concepts and connect them to what has been learned before. For the brain, the most effective approach is to focus on one thing at a time. Therefore, frequent context switches and multitasking—temptations often introduced by AI tools—tax working memory. Jääskeläinen concluded with a clear message: study well, but also sleep well, because without sleep learning remains incomplete.

What are we learning when we build something?

University lecturer Meri Kuikka and Master's student Ira Lahtinen described what and how we learn through building. Construction-based learning—for example, “Toolbox Full of Legos”—enables multidisciplinary collaboration, idea testing, and the realization that problems can have multiple correct answers. Encountering disagreements, experimenting boldly, and rapidly developing prototypes help students reflect: What did I assume? What did I do? What did I learn? What will I do next? As the speakers noted, “being wrong is cheap”; learning begins with mistakes. When AI offers direct answers, the value of construction-based learning lies precisely in testing, making thinking visible, and learning together. To support considering different perspectives, the speakers also suggested trying the Six Thinking Hats framework.

Who's actually learning here, you or the AI?

University lecturer Antti Lähtevänoja emphasized that even in the age of AI, learning requires effort. Generative language models are based on probabilities and do not understand the big picture. Therefore, we must learn to use them correctly: for checking and sparring, not for ready-made answers or doing the thinking for us. Lähtevänoja urged teachers to protect the “friction” of learning—designing teaching and schedules so that AI cannot do the thinking on students’ behalf. AI does not replace thinking: if you outsource thinking to AI, no learning takes place.

Learning now: A student’s perspective

Bachelor’s student in Mathematics and Systems Sciences Konsta Kiirikki outlined the differences in students’ AI and study skills. While some students don’t use AI at all, others build their own tools to complete tasks on their behalf. The work of a skilled student can look similar to AI output. Kiirikki asked whether the university could be a place to learn both field-specific AI competencies and study skills. In practice, this could mean course-specific learning objectives for AI use, ensuring equal access to AI tools, and assignments that assess understanding and the learning process. It is important to make AI use transparent and discuss it openly with students.

From Instruction to Student Agency

University lecturer Tomi Lauri emphasized the importance of agency. In the bachelor’s-level film education, students are treated as filmmakers from day one, which helps them learn to build knowledge together and take responsibility for their own learning. The teacher’s role is to act as a facilitator. In the age of AI, we need ways to assess learning rather than just outputs, make learning processes visible, and discuss authorship and the role of tools. Agency, collaboration, good feedback practices, taking responsibility, and critical thinking remain essential parts of university teaching.

How do we respond to the challenges of AI? 

After the presentations, participants discussed their takeaways in small groups. The discussions considered how learning is changing and what challenges AI has introduced to teaching. The groups emphasized the need to develop AI competencies for both students and teachers. Participants called for transparency in AI use and for teaching methods that ensure students’ voices are heard. Proposed solutions to assessment challenges included oral presentations, alternative ways of completing course assignments, and step-by-step documented work in which students also disclose the prompts they give to AI. Tips for using AI should be shared collectively, and students’ responsibility and agency can be strengthened by making clear the consequences of AI misuse for learning. Even in the age of AI, we are ultimately facing fundamental questions: how do we teach students to learn, and what is learning, really?

When AI produces ready-made answers, the value of university teaching, based on the morning’s perspectives, lies in making thinking visible, fostering collaboration, and strengthening students’ agency. AI is just one tool for learning – learning still requires a thinking human being.

The presentations (recorded on video) will be added to the news article later. The next Marsio Presents event will be held in spring 2027.