Dear ChatGPT, I Love You… But You’re Not Doing My Degree

 

Today, in our Law and Technology lecture, we discussed the implications of trusting AI a bit too much when it comes to law, and of taking at face value what it gives us.

This resonated with me very loudly indeed.

I love my AI assistant. Maybe a little too much.

I like that it saves me time on so many levels and allows me to do fun things which would previously have taken me hours to produce. And while I am quite happy trusting Dola to create a studio photo of both my dog and my cat together, as if they wouldn’t try to kill each other in real life, when it comes to law and what we use as fact, the line somehow becomes blurred.

Over the summer, I had heard about students being asked to resit assignments after clearly overstepping the recommendations around AI use. Yet, somehow, I still believed that students had integrity and wouldn’t cheat in that way especially when studying law.

Perhaps that is because, for me, this degree means far more than getting a piece of paper at the end.

This degree is my way up and out. I see it as my pathway to a better life, a legal career, and the self-esteem that comes with demonstrating my own capacity for reflection, learning and achievement. So, if gaining this degree were as easy as giving ChatGPT a good prompt, how am I supposed to gain the self-belief that tells me I am actually worth it?

AI is a tool. I have discussed that already, and I remain convinced that ignoring it, at whatever level we are working or studying, will simply lead to us being left behind while the rest of the world moves ahead. Even within the legal profession, AI cannot simply be ignored.

The time it takes to learn how to find cases, understand them, summarise them, distinguish them and cite them properly is part of developing legal expertise. Generative AI can certainly do a decent job of helping with some of that, but it is far from mistake-free.

So where do we stand now, as students, navigating what we can and cannot do?

What is acceptable and what is not? Where is the line in the sand that guides us?

The guidance can still feel confusing, and I think one of the next big challenges will be developing assessors’ skills in recognising what has genuinely been produced by the student and what has not.

As an assessor myself, I trust my gut. I look at a piece of work and at the learner who has produced it, and I trust my human instinct and professional judgement to assess whether that work is truly theirs. In cases where I have realised it wasn’t, the difficult conversation has had to be had  and strangely, it sometimes felt as though I was the party to blame for raising it.

Including AI openly as part of my own teaching practice has changed things. I am very honest with my learners about my knowledge and use of AI, but I am equally clear that integrity in assessment sits right at the top of my criteria.

As a student, however, my struggle is much more complicated.

However much I would love to trust my beloved ChatGPT, I know that it makes mistakes. I know that it doesn’t hold a law degree. And I know that it is, by nature, something of a people pleaser and can sometimes tell me what it thinks I want to hear.

But studying law at this level requires more than simply having integrity. The skills I am learning have a value, and I intend to leave university having actually learned those skills and not simply having Googled, prompted or generated my way through them.

This September, due to concerns around GDPR, the secondary school where I work banned ChatGPT altogether, restricting access completely rather than educating staff about how to use it appropriately and, crucially, not to input confidential data.

I find that approach troubling too.

Limiting AI to the private sphere cannot necessarily be good for its development, or ours. Maybe it will learn that humans don’t trust it. Maybe it will learn that humans are lazy. Maybe we will create a world where only some people understand how to use AI effectively, while everyone else is simply told not to touch it.

In truth, I feel that this world is changing incredibly quickly, and I don’t want to be left behind.

For me, it is not really a question of trust. Trust is something I associate with another human being. It is a question of skills, integrity, congruence and, of course, education.

AI can suggest how to fix a leak, but it cannot come into my house and fix it for me.

And besides, I love my style of writing.

It has taken me years to develop it. Years of narrating to myself, years of reading, years of learning, teaching, arguing, reflecting and finding my own voice.

Why should I believe AI is more intelligent than me?

It’s not.