AI-models better than the experts

In the newsletter Curie, April 3, three researchers describe their work and how they use AI in their work.

Thomas Schön, Beijerprofessor in artificiell intelligens at Uppsala university and scientific leader of the five-year project AI4Research, means that an important piece of the puzzle in all AI-research is tho show what the technic do to us as people and society. He says:

– We must remember that this technology doesn’t think, it counts. Therefore we must make sure that the data is used in the right way, and that technology is built according to human values.

Elisabeth Epstein is professor/Senior Physician, Department of Clinical Science and Education, Södersjukhuset at Karolinska Institutet. Her research team have developed an AI model with which they can distinguish benign tumors from cancer. The team expresses that even if AI is more accurate in assessing cancer, we must not rely on it completely.

Artificial Intelligence – Digital Essentials

“Artificial Intelligence” by the University of Queensland (UQ), provides an overview of AI, including its types and applications. The module discusses the use of AI at UQ, including guidelines for students and staff, delves into the legal, ethical, and social implications of AI, and offers practical examples of AI applications. The module also guides readers on how to prompt AI tools, evaluate AI-generated content, and properly cite and publish with AI.

Legally secure examination and Artificial Intelligence

“As AI tools are increasingly used, our educational institutions and the way we conduct examinations are being challenged. At the same time, we need to educate our students so that they can use these new tools in the best possible way. In this special issue of the journal Högskolepedagogisk debatt, the relationship between legally secure examinations and AI is highlighted. In the introduction, Anders Eklöf & Lars-Erik Nilsson from HKR write, ‘There are no given answers, so it is set for us to become the kind of knowledge organizations we are meant to be – inquisitive, unpretentious, and open to debate on how AI applications can be used in almost all aspects of our operations in a defensible manner.'”

Notice that the pdf it´s only written in Swedish! Download and read Högskolepedagogisk debatt: Rättssäker examination och artificiell intelligens. Johan Landgren, Anders Eklöf & Lars-Erik Nilsson (Red.) (2024) Kristianstad University Press

Posted in _AI