The AI debate – inevitability or threat?

Following the so-called AI ‘boom’ of 2022, it seems as though every app and website comes with an AI feature, and every conversation mentions at least in passing some of the most powerful software in the world. But student perspectives seem to remain mixed. In pursuit of gaining a better understanding of the sentiments towards AI of the Warwick student body, I created and distributed an anonymous Microsoft Form, focused primarily on the concept of AI denialism. It should be noted that, likely due to the nature of the circles I am a part of, the majority of responses were from students studying for a humanities degree.

71% of Warwick students said they do not use AI in most of their assignments, with a similar number believing that AI is not the best way to get work done

I attempted to offset this by having a STEM-studying friend share the poll on his Computer Science society online group chat, and while this greatly increased the number of individuals studying STEM (and particularly, by my design, computer science), it remained that 68% of the data I collected still came from humanities students. That said, I had included the question of degree on the form with a smaller investigation in mind, having assumed that humanities students would be less AI-positive than STEM students. I was a little surprised to learn that this wasn’t the case, and that attitudes to AI did not vary consistently between the two subject areas, and the schools of thought that I had anticipated did not actually emerge.

71% of Warwick students said they do not use AI in most of their assignments, with a similar number believing that AI is not the best way to get work done. That said, far more people (33%) became happier with AI use in studying, but not for doing their work for them. University policy promotes ethical use of AI, with an emphasis on academic integrity, including not submitting AI-generated output as one’s own work. The wording of my poll questions was vague in some respects, as my principal goal was to gain a broader overview of student use of and outlooks towards AI, while also being aware of the possible aversion some might have to admitting their AI use, despite the anonymous nature of the poll.

Despite the vast majority (66%) claiming that they try to resist AI, there looks to be a sense of acceptance (a similar kind to that in the five stages of grief) of the extent to which AI is intertwined with the future

For similar reasons, my poll also did not delve particularly into the varieties of AI, such as the differences between agentic models that have been in use since the 1960s, typically used for decision-making and data analysis, contrasting the content creation focus of generative AI models. Generative AI (GenAI) has become something of a buzzword in recent months, having taken leaps and bounds beyond the agentic AI that seems to have evaded the public eye, and the initialism itself, it appears, has almost become a shorthand for generative artificial intelligence. For this reason, my investigation focuses on this branch of AI specifically.

When I asked whether people felt positive about the future of AI, only 8% answered that they did. Compared to the 56% who agreed to a small or large extent that AI is a part of the future, whether we like it or not, the outlook emerges rather bleak. Despite the vast majority (66%) claiming that they try to resist AI, there looks to be a sense of acceptance (a similar kind to that in the five stages of grief) of the extent to which AI is intertwined with the future. Countless students noted that AI denialism is an ultimately disadvantageous stance, resulting in falling behind the curve in the workplace.

On AI in the job market, results were slightly more weighted towards student concern over AI replacing them in their (future) careers, and while the majority argued that AI would not enhance their jobs, a fairly sized 24% believed that AI would enhance their jobs, with a further 14% sitting on the fence. The relationship between AI and the job market is currently unstable and unclear, but the fear of being left behind prevails, bolstered, I would imagine, by the uncertainty of the future.

Indeed, when the issue of whether AI was a bubble that would eventually burst was raised, responses were split almost evenly between agreement and disagreement. Most of those who believed that this was the case also stated that they followed AI in the news, while those who did not believe AI was a bubble that would burst were more likely to not stay up to date with AI in the news. Student attitudes to AI can also be explored through weighing up what are seen as the benefits and flaws of its use. The most popular pro of AI was data analytics and predictions, perhaps signifying an attachment to the familiarity of the more accepted forms of AI, which are perhaps so ingrained that they do not have the same associations as the newer GenAI models.

Advancements in medicine, understandably, came in close second in terms of popular benefits of AI. In spite of the many flaws of AI discussed below, one would be hard-pressed to dispute the value of more advanced technology in medical discoveries, particularly in the ongoing battle against cancer. Interestingly, however, the least popular gain of AI use was economic growth, with 20% of students citing this as a reason to value AI. This is most likely tied to the aforementioned fear of job loss, which 80% of students testified to be a potential flaw of AI expansion.

However, the most popular flaw of AI use was reduced human connection, a consequence, I imagine, of the chronic loneliness reported in our generation. Interacting with AI bots, for example, rather than customer service workers, having your work checked by an AI tool, rather than your boss, might contribute to a sense of isolation as we replace human interaction in the name of productivity.

This AI generator offers a service that generates bedtime stories, complete with the tagline ‘Boost your productivity and focus on what really matters’. But what really matters? What are we saving all this time for? If we even need to offload the task of making up a bedtime story for our children to artificial intelligence, what do we do instead? What is so important that we stop wasting our time on thinking and creating ourselves? What do we put our time into instead? What really, really matters?

Several students, in the customisable ‘other’ box in the ‘flaws of AI’ question on the form, highlighted an area that I had purposefully excluded in the interest of addressing the brief for this article: the relationship between art and AI. To be pithy, those who raised the issue of AI art were heavily condemnatory of its existence and use, and I myself must agree. Such a matter is fairly adjacent to another flaw of AI use: human dependency, and a reliance on artificial intelligence as a result of an erosion of natural intelligence, which is what I consider to be the greatest detriment to the growth of AI.

My intention is not to write a fearmongering article which may as well be titled, in the words of one poll responder, ‘Terminator’, but that does not mean my viewpoint is not rather grim. MIT recently carried out a study revealing a side effect of AI use dubbed ‘cognitive atrophy’. Will human desire for greatness outrun itself? Paracelsus, a Swiss physician in the early 16th century, wrote of human scientific development that ‘We shall be like Gods. Natural magic will make it possible to see beyond the mountains, to divine the future, to cure all diseases, to make gold, and even to duplicate God’s greatest miracle – the creation of man himself. The description is obviously familiar: AI is lauded for its ability to predict, to advance healthcare, to boost the economy – and is frequently oddly pseudo-humanised.

Supreme caution, and constant questioning are necessary for an engagement with artificial intelligence

But will humans ourselves, sitting in the superiority of critical thinking and opposable thumbs, be undone by our own hubris? I return to the question of why we would need AI to generate art for us without life. Why would we want it? Why should we want AI to do our thinking for us, if that is such a joy of being human? Its positive power in some areas does not justify its presence in others.

Supreme caution and constant questioning are necessary for an engagement with artificial intelligence. Even if it is a bubble that may burst, this is by no means a certainty, and taking it as an inevitability, rather than falling into the mindset of AI denialism, is perhaps the best way to navigate the changing landscape.