Does ChatGPT make students lazy thinkers? — Insights into a sensible approach to AI at the University of Bern

In brief:

Artificial intelligence (AI) is developing at a rapid pace and is increasingly able to take over human work. As a society, we are therefore faced with the question of how we want to shape our collaboration with artificial intelligence. In a conversation, philosopher of science Prof. Claus Beisbart provides insight into the integration of AI at the University of Bern.

The future of the lazy

The UniBE Foundation focuses on supporting research that is forward-looking and innovative, which is why AI is a recurring topic for us. The areas in which AI can work are steadily expanding. Today, an AI such as ChatGPT can already take over creative aspects of human work – text production, for example. It seems that AI systems are increasingly relieving people of thinking and solving problems. So perhaps, thanks to AI, we will soon be able to spend more time on our hobbies. But what if widespread use of AI makes us lazy thinkers? For a think tank like the University of Bern, such a development would be quite dangerous.

In my conversation with Claus Beisbart, Professor of Philosophy of Science at the Institute of Philosophy, I want to find out what consequences the use of AI is already having at the university today and what social challenges we will still have to face in the future.

Interview with Prof. Claus Beisbart #

UniBE Foundation: Professor Beisbart, what led you to make AI and ethics one of your research priorities?

At the moment, many paths lead to artificial intelligence. In the philosophy of science, I had already worked on computer-based methods. The focus was mainly on computer simulations, which are used, for example, in climate science to make predictions. Currently, however, computer simulations are increasingly being replaced or supplemented by so-called ‘machine learning’. There is a real hype around this ‘machine learning’ at the moment. I am interested in what happens to the sciences when such methods, which are largely ‘black boxes’ for us, are applied. Ethics quickly becomes relevant here too. For example, one ethical question is what we would need to know about an AI application in order to be able to use it in medicine.

What practical ways do you see of bringing inclusive and transparent AI into research?

There are very different approaches. A major problem for universities is that ChatGPT and other large models come from industry. We did not make them ourselves and can only examine them after the fact, just as all users can. The university should take the lead here and develop models itself. Another approach is explainable AI (XAI). This means that the AI is built from the outset in such a way that it is easier to understand. Approaches focusing on interpretability try to make algorithms understandable after the fact through close examination. Such methods would need to be developed further at the university so that they can then also be used in practice.

And what can be done in teaching?

Basically, as a university, we contribute to educating people about AI. It is important that people know exactly how AI works, what the dangers but also the opportunities are, so that they can handle it with care. We have an important educational mission here. In this context, we absolutely must use AI. Shutting ourselves off from it achieves nothing. It is about finding a responsible way of dealing with it.

You have developed a digital module on ethics and AI for students. What motivated you to do so?

The Vice-Rectorate for Teaching had the great idea of creating an online module “Skills for the (digital) future” for students of all subjects. I was happy to contribute on ethics. Society must create the framework for embedding AI in our lives in a compatible way. It is important that this is not driven solely by corporations pursuing their own policies. I think we humans need to consider where and how we want to use AI. Ultimately, that is exactly what the module is about for me: encouraging people to think about AI for themselves.

Click here for the teaser of the online module “Ethics and Digitalisation”.

Since its launch just over a year ago, ChatGPT has become a constant companion in everyday university life. What challenges do you see in dealing with this AI in a university context?

We have already talked about the opacity of AI. ChatGPT is also not trained to tell the truth or cite sources. That is very inconvenient for academic work. So you have to think very carefully about what you use ChatGPT for. I see ChatGPT as an attempt to form a kind of average of everything that is said in the training data from the internet. So if I am interested in what is generally said about a topic, ChatGPT is the right place to go. But if I am interested in the truth, or in the reasoning and sources, then ChatGPT becomes difficult.

How have you experienced students’ use of ChatGPT over the past year?

This semester, I taught our methods course, in which philosophical writing is practised, for the first time since ChatGPT was introduced. For smaller tasks, I allowed the students to use ChatGPT for help. However, the enthusiasm was limited. In general, I have the impression that our philosophy students like to write themselves. They recognise that they need to learn to write and are also interested in doing so. In philosophy in particular, linguistic expression and thinking are closely interlinked.

Papers and essays are susceptible to being written with ChatGPT. One option would be to introduce other formats, such as oral examinations, for assessment. What do you think of that?

I do see a certain problem with ChatGPT. It is getting better and better, and there are also other language models that can be used. With clever prompting, it will probably soon be possible to get ever closer to a good seminar paper. It is difficult to distinguish such a paper from ‘genuine’ seminar papers or essays. There are also ways of editing the text afterwards and inserting a few spelling mistakes so that it looks more like a student’s work.

Even though there is a risk of cheating, I would not want to go so far as to abolish the seminar paper. In philosophy, seminar papers are very important. Students engage in depth with a topic over a longer period of time. Ultimately, it is about intellectual development, which for me is part of personal development. Students question their assumptions, introduce new concepts and refine their arguments. This process is relevant for their own reflection. That cannot be demanded in the same way in an oral examination. What can be done, however, is to discuss a seminar paper afterwards. This allows me to see how well the student has understood the subject and whether the submitted work really is their own. Such a conversation is also pedagogically valuable for other reasons – for example, to clear up misunderstandings that may have arisen during marking.

Digitalisation requires humans and machines to work together. What opportunities does this offer?

The big opportunity is that AI takes over work that we do not like doing. So AI could take over uninteresting or boring tasks. There is also the possibility that AI will do this work better than we do. In medicine, for example, AI sometimes provides better diagnoses than humans – although these are very specific areas of work.

And what are the risks?

I work with computer-based methods. My team and I have a research project that tries, to some extent, to implement moral and philosophical reflection on the computer, which can be seen as a rudimentary form of AI. However, this is not the ‘machine learning’ that is currently so hyped. Nevertheless, it is also a kind of artificial intelligence, as human thought processes are replicated by a computer. Moral thinking is simulated to some extent. I also have some ideas about how I could use ChatGPT in research, but I have not yet put them into practice.

Do you think researchers will one day become unnecessary because of AI?

No, I don’t think so. Although it has to be said that AI is very good when it comes to individual tasks. It is always specialised tasks that AI takes over, and where it can be trained to be better than a human. What AI still lacks, however, is general intelligence. Nor can AI easily take over the decision about a relevant research topic and how I use my time and resources. We definitely want human science, not science that runs automatically and researches something over which we no longer have any influence. As things stand, humans are also needed for many intellectual tasks, for example to make cross-connections. So for the foreseeable future, humans will still be needed.

After half an hour, I say goodbye to Professor Beisbart. The conversation showed me that AI certainly opens up new and interesting ways for us, at the university and in society, to deal creatively with knowledge. But it also made me think and showed me that, as a society, we play a decisive role in shaping the use of AI in everyday life as well as at the university. We should therefore not shy away from engaging with it.