How is AI Transforming the Learning Process in Academia?
Artificial Intelligence (AI) is revolutionising education, from primary schools to PhD programmes. But is it really AI that’s upsetting the PhD education system? Or is the technology mainly exposing deeper, underlying issues? Vicent Botella-Soler questions whether current PhD training truly fosters critical thinking and suggests that AI offers an opportunity for improvement in a system too focused on results. Let’s just compare it to… a surprise visit from your in-laws?
Disrupted: AI and the PhD
AI tools seems to have broken something in doctoral education – and in education at large, no doubt – however confused we might still be as to the nature and extent of the damage. What appears clear is that AI tools are set to fundamentally change the way research is practiced: from literature reviews and data analysis to coding and experimental design.
One pain point (though I believe it a blessing) is that these new tools challenge the way we currently evaluate researchers. How can we fairly credit and assess research outcomes when AI assists in writing papers or grant proposals? Whose idea was “that”? GenAI has put a spanner in the works of our highly competitive, metrics-driven academic system. But let’s not forget that our system of metrics and evaluation was already in crisis before the arrival of AI. We have been obsessed with productivity and “excellence” for years, using proxy-metrics to measure and rank everything from papers to researchers to institutions, often to the detriment of quality and relevance. As Goodhart’s law predicted, people have found ever more inventive ways to game the metrics-driven academic system (including scientific fraud, in the worst cases). AI tools have just exacerbated these problems.
In my opinion, the crux lies somewhere else: how will AI affect the learning process of our students? Learning is often about doing – thinking, reading, researching, summarising, etc. If AI can do many of these tasks for us, will we learn as much? Or will we learn differently? Can some intelligent use of AI enhance the learning process? No doubt. However, demands for increased productivity are often in direct conflict with the idea of learning as a process. We are told AI tools can “save time for more important things”, but what are these more important things? If learning through doing is essential, perhaps we need to continue doing, even if AI could do it better and faster. And what will be the real effects of AI tools on our already strained productivity expectations? What will this do to the current mental health crisis in academia?
Even if we agree that the PhD learning process is vital, academia is still stuck evaluating outputs, such as theses, grants, patents, publications. Standing out as a refreshing counterexample, some institutions, like KU Leuven, are already adapting the criteria of the PhD to be truer to the idea of learning as a process, and not placing so much emphasis on the concrete outputs of the PhD research.
AI forces us to reexamine fundamental questions:
- What is a PhD for?
- What should students learn from this experience, regardless of discipline or research topic?
- How can we ensure the learning process during the PhD in the AI era?
In a recent meeting of doctoral education professionals in Zürich
(Switzerland), several speakers highlighted the importance of developing
critical thinking habits. Although there has always been an implicit
expectation that PhD students learn to evaluate ideas and statements
critically, we might benefit from teaching critical thinking explicitly.
One speaker emphasised the need to “put the philosophy back into the
Doctorate of Philosophy!”, with the same enthusiasm one would shout “to
the barricades!”. I fully agree. We should teach philosophy of science
and epistemology during the PhD; explicitly encouraging students to question their own thoughts and knowledge, and to continuously reflect on what it means to “know” something and the different ways we can access knowledge.
Let’s not lose sight of the fact that, at its core, the AI discussion is nothing but a reprise of a much older conversation about the role of academia in society, and its methods. As I see it, the impact of AI on academia is, so far, like a surprise visit from your in-laws: the house was already dirty and in disarray, their arrival just makes the mess more obvious and painful.
References
- Moore, S., Neylon, C., Paul Eve, M. et al. (2017): “Excellence R Us: university research and the fetishisation of excellence”, Palgrave Communications, 3, 16105. https://doi.org/10.1057/palcom....
- Iagoli, M. (2020): “Fraud by numbers: Metrics and the New Academic Misconduct”, in Los Angeles Review of Books, 7 September. https://lareviewofbooks.org/ar....
- Muller, J.Z. (2018): “The Tyranny of Metrics”, Princeton University Press
- Hall, S (2023): “A mental-health crisis is gripping science — toxic research culture is to blame”, Nature 617, 666-668. doi: https://doi.org/10.1038/d41586...
AI for researchers
Introduction to AI and Machine Learning for Biomedical Research
training course - online - KU Leuven, VUB, UHasselt, UGent, VAIA
Generative AI for Professionals in Higher Education
learning trajectory - VAIA & VLIR
Data Visualization
course - Bruges - PUC KU Leuven Continue, VAIA
Vicent Botella-Soler
With a background in physics and computational neuroscience, Vicent has worked in data science and machine learning in the private sector for many years. He currently works as a freelance consultant and trainer for academic institutions, on topics related to academic well-being, decision-making, and critical thinking.
Related courses
HPC Café
Seminarie - Online, Karlsruhe - KIT, HammerHAI, Training geselecteerd in de Belgische AI Factory Antenna (AIFA)
