OPINION: Should students miss out on learning how to create new knowledge?

This is an opinion piece, the views expressed in the column are the writer’s own. 

Maja Horst and Niels Lehmann finish their reply to Mary Hilsons opinion piece in Omnibus with a sports metaphor. Following this, I would like to join Mary Hilson as her partner in a mixed doubles match.

Reply: Maybe we don’t disagree that much – just on one point

The dean and vice-dean claim in their feature article in Politiken that if you’re able to make an effort, GAI delivers high quality. The article reads that "a relatively weak student (can) produce an impressive Master’s thesis in less than a week”. They further claim that archaeology "20 years ago, was a craft based on excavations and archival studies.", while today, simulations and data models are being made.

OPINION: "I don't accept the premise that implementation of AI is predetermined and unstoppable"

The claims are presented without evidence and even demonstrate a lack of knowledge about the disciplines they are set to lead.

Everyone can recognise that when new academic regulations are written, we must define the learning outcomes for knowledge, skills and competencies for each module of the course. In the article, Horst and Lehmann focus on competencies, but forget the rest.

Where does all this 'knowledge' come from, which GAI has access to? Yes, big tech companies have taken a lot of information from whatever they could access and stored it in huge data centres that use a lot of energy. And we must then continue to feed and become addicted to this monster. The next step is to throw out our dear libraries and burn the books. Then we have reached the "point of no return".

The art of creating new knowledge

How then is "new knowledge" obtained? Not from GAI at least. GAI is a babbling parrot-like technology that "hallucinates". GAI is not able to identify new research perspectives or gaps in a field that is based on empirical data. It has no concept of unopened archives, unprocessed findings, or new methods of investigation. Is GAI able to interpret historical texts? Is GAI able to interpret documentation from archaeological excavations? Is GAI able to analyse material from new theoretical perspectives? I doubt it very strongly. In short: GAI is not a technology developed to promote creativity, originality, intelligence and other valuable human qualities in the scientific process.

In addition, there are skills. It’s a misconception to perceive excavations as a craft, even though we use our hands. Excavations must be learned as a reflexive process, where observations and expectations (knowledge) are continuously in dialogue. Without ones own knowledge at the moment of the study, no dialogue. And in a similar way, reading and interpreting texts is a reflexive process.

Of course, new technological opportunities must be made use of in the future. But just not uncritically. If a new technology can solve manual or other tasks, this relationship must be utilised to promote the quality of the scientific process. This perspective is unfortunately completely absent from Horst and Lehmann. They are focused on the fact that the degree programmes serve the purpose of "getting good lives and working lives in a digitally transformed society”. This is a commendable goal, but the programmes serve (first and foremost) to improve one's skills in a scientific discipline with the purpose of acquiring new knowledge. This art must be learned, it must be embodied in the student himself. And the students shouldn’t miss out on this experience.

This text is machine translated and post-edited by Lisa Enevoldsen.