Image: Warwick Media Library

Does the need for AI reflect poor teaching standards at Warwick?

A survey by The Tab found the University of Warwick had recorded more ChatGPT site visits in December and January than any other Russell Group university. During that period, people at Warwick made over 850,000 visits to the artificial intelligence (AI) website — five times more than the University of Newcastle, which was the second-biggest user.

“Mainly [I use it] to understand difficult theory”, said a second-year Liberal Arts student. “I ask it to give a summary of x text. It’s not always accurate, but it’s close enough for me to understand”.

She wasn’t the only one who used ChatGPT as a resource to break down and explain complex ideas, as well as to ask questions about academic subjects. The chatbot, which was developed and launched by OpenAI last November, can provide detailed written responses when prompted by users. Students said they used it to find answers quickly, rather than using search engines, as ChatGPT filters through information online and only supplies material relevant to user prompts.

“I could easily see this [replacing] search engines in the future.”

– third-year student

A third-year Computer Science student explained the chatbot was more beneficial than trawling through endless search results to find what they were looking for.

“What I love about using ChatGPT in my work is that it will give you specifically what you want, instead of you traversing through Google pages.”

The AI chatbot experienced a boom in popularity after it became available to the public and many people revered the bot as being the future of technology, since it was able to generate pages of realistic writing. ChatGPT even attracted a multibillion-dollar investment from Microsoft.

However, with the success of the chatbot came fears that it could be exploited by students to write assignments, essays, and even exams. With the AI constantly learning from its users, ChatGPT can easily mimic the writing style of a real person and use the information it has consumed online to write about every topic imaginable. For example, the chatbot has passed several law and business exams with incredibly high scores. Many schools and universities (including Oxford and Cambridge) have already banned the use of AI bots in exams.

There were more than one million visits to ChatGPT by students during the winter exam period.

The University of Warwick said it would be taking an “active approach” to tackling the problem, following the high usage of AI chatbots during exams, and that the design of exams would be reviewed. Warwick said future exams would use “detection techniques showing where AI has been used inappropriately” to prevent this new method of cheating.

There were more than one million visits to ChatGPT by students during the winter exam period.

Professor Till Bretschneider, Director of Undergraduate Studies at Warwick’s Computer Science Department, said: “There are some active discussions going on, but as it affects the University and the sector as a whole, it is probably too early for the department to comment.”

With summer exams currently ongoing, it is unclear what the university is doing to prevent students from using AI assistance in their exams and final essays. There seem to be no concrete plans to restrict cheating beyond methods already employed.

I don’t use it directly in my essays because I don’t think it’d do very well.

Despite the widespread panic about cheating at UK universities, the high usage of AI among students indicates a different problem entirely. Most students are reluctant to submit an essay entirely written by ChatGPT, either because they consider it to be cheating or they fear it wouldn’t be as analytical as one written by a human. Although the chatbot can generate plenty of information, using material it finds online, it can’t form an argument or provide any decent analysis. With this in mind, why was the site accessed so often over university Wi-Fi?

After investigating what the Warwick community used ChatGPT for, The Boar found it may suggest an underlying problem of poor teaching rather than increasing dependence on technology. Speaking to students, a common pattern was unearthed: it had become a tool for explaining difficult concepts, summarising readings, and otherwise filling the gaps left by staff.

Whether it’s due to strikes, inaccessible teaching, or the structure of modules, many students feel they don’t understand course material and can be confused by difficult concepts. This can make it difficult to complete assignments and nobody wants to risk a poor grade when they’re under pressure to succeed — from the university, their parents, and the world of work.

Although university staff have largely opposed the use of AI, students can’t be faulted for using the chatbot to explain concepts if teaching staff have failed to do so themselves.

“It’s not perfect but it has given me ideas or helped me come up with them.”

ChatGPT has also become a source of ideas and has been utilised by students who need inspiration for their work.

“I have been using it to get ideas about what else I could write about in my report that I have due, and it’s great for brainstorming other ideas I haven’t considered.”

It could be argued this is reducing students’ ability to think creatively, by allowing AI to generate ideas for them, and could lead to future generations failing to develop skills necessary for the world outside of education. Young people are frequently accused of being lazy, entitled, and self-obsessed, and the development of technology which allows them to ‘cut corners’ is used as cannon fodder for this argument.

To some extent, people are becoming more dependent on technology, as it becomes a more crucial part of their lives than ever before, and this may result in the way they work changing dramatically.

But it’s more likely ChatGPT has become a crutch for students who are overworked, stressed, and aren’t finding support anywhere else. Seminars being cut due to striking means fewer opportunities to discuss and brainstorm ideas with other students, leaving people feeling isolated when they have to do so. AI replaces the seminar discussions that so many people have missed out on this year by collecting concepts and generating new ones.

Similarly, students struggle with writing essays and feel they are unsupported by the university when doing so. In this case, being able to ask AI questions any time of the day and knowing they’ll receive an immediate answer can reduce stress.

Contacted for comment, Will Brewer, Warwick SU President, said

At this current time the SU are not involved in any formal discussions at the central University level regarding ChatGPT, but we are keen to engage on this topic in future discussions, exploring how the curriculum may change over time to reflect the advancements in AI whilst ensuring the fair assessment of all students.

Although the widespread panic about AI is understandable — it can now learn, write, and create like a human being — humans are always frightened by the advent of new technology and this fear can be traced throughout history. For example, electricity, elevators, and automobiles were all perceived as terrifying by many at the time of their invention, but are now commonplace around the world. Today, most people would consider fearing electricity to be an irrational fear to have, despite it being seen as a supernatural concept by their ancestors.

ChatGPT has also become a source of ideas and has been utilised by students who need inspiration for their work.

While the idea of ChatGPT becoming part of daily life might seem a disturbing concept, it doesn’t seem far-fetched when considering how many people already live with virtual assistants, such as Alexa and Echo Dot. In many ways, it represents the natural evolution of search engines and how they’ve integrated with built-in, voice-controlled personal assistants like Siri. The concept of AI chatbots isn’t exactly new, either; with the first one being made in 1966 and many others being programmed since, ChatGPT is only the latest in a long line.

As with many trending topics, ChatGPT’s popularity may be coming to an end already. Now, the Snapchat AI is the latest bot everyone is talking about and it currently faces the same mixed reactions its predecessor did.

Critics may rush to blame modern technology for producing a generation of ‘lazy’ students, but the issue at hand here isn’t with AI. The problem is that both students and staff at universities are overworked and undervalued, with tools like ChatGPT providing them with the support they desperately need.

Comments

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.