‘Eminently sensible’ is how Professor Michael Madden, Head of the School of Computer Science at University of Galway, described our GenAI policy on the MBA final year project module.

As yet, to my knowledge at least, what we do is somewhat unusual for our area. It was good to have the feedback.

Three weeks ago, Professor Madden came to speak to our second-year class.

His talk began with the modest birth and rapid growth of AI and specifically Large Language Models (LLMs). Detailing the current state of play, he distilled the sense from the hype, and the legal wranglings between publishers and providers. The talk contrasted the power of the algorithms with the biases in the database, the hallucinations, the fake answers.

Having seen all this grow from the early days of machine learning research, today Michael has an interesting take on how scholars and students might use AI.

In the Q&A part of the lecture, a student asked his views on using ChatGPT to write professional marketing blog posts, a handy feature for sure. The way he sees it, he told us, if a piece is not important enough for the author to actually write themselves, why make someone read it? If the author does not feel sufficiently compelled by what it is they have to say, to put their own pen to paper and convey a thought, is it really worth communicating?

This gave me pause- to think of creation of text as an act of responsibility—a kind of contract between writer and intended reader -- if something is put out for people to read, are we wasting their precious time by substituting that jumpy little cursor prompt for our own conscious thought?

Of course, he added, LinkedIn Posts, short summaries, those kinds of things are different—the tech can be useful in such cases*.

But another thing to ask, is why do we write at all; what does it do for us? He writes to figure out what he thinks in the first place, he told us, to see the arguments for both sides. To sketch out, to take detours, to watch unanticipated problems and paradoxes materialize on the page. Writing is a process, a musing, a meaning-making.

I thought of Joan Didion’s Why I Write. Encountering any scene, she tells us, she writes to discover " what I’m looking at, what I see and what it means...” She writes “to find out what I’m thinking".

Another question, Michael went on, is we need to figure out why it is we read. To absorb and consider, or to crunch pieces of text to turn out summaries?

Existential considerations of why we read and write were not what I was expecting from this lecture [ostensibly on ‘AI in Research’], but a return to first principles is always welcome.

And as to our GenAI policy on the MBA project? We ask students to use GenAI to brainstorm ideas, to second-opinion data analysis, to suggest overlooked topics, and to reflect on it all. This machine should be approached as one would an unruly intern: eager and friendly but basically unreliable.

A solid approach, was Professor Madden’s appraisal, especially given what we still have yet to learn about this this emerging tech.

*Not, of course, in this case.

Kate Kenny

First published 4th October 2024 on LinkedIn