When David Bowie sat down on Newsnight with Jeremy Paxman in 1999, his comments on the advent of the internet struck an eerily prophetic tone. “I don’t think we’ve even seen the tip of the iceberg,” he argued to an ever-sceptical Paxman. “I think the potential of what the internet is going to do to society, both good and bad, is unimaginable. I think we’re actually on the cusp of something exhilarating and terrifying.” Resisting Paxman’s insistence that it was just “a tool”, Bowie asserted that it was in fact “an alien life form!” The angst, the exhilaration; the turn of the century was marked by this heady contradiction of emotions as the public anticipated a digital future with both eagerness and cynicism.
Although Bowie would live long enough to see his predictions justified in manifold ways, he was precluded from witnessing the birth of a new age of technological advancement. Dying from cancer in 2016, just two years before OpenAI released their first GPT (Generative Pre-Trained Transformer) models, he would never live to see the internet become a conduit for perhaps the most “alien life form” we have at our disposal: generative AI (Artificial Intelligence).
Such is the ubiquity of generative AI in our everyday lives, it is difficult now to imagine a time when we didn’t have to query if novels were written by human beings or whether images and videos could be taken as fact; a time before harmful content could be generated by a simple prompt and the click of a button. Equally, it is difficult now to imagine a world in which tedious admin tasks couldn’t be immediately resolved, or where hyperspecific research questions couldn’t be answered in mere seconds. Even as I do research for this article, it is not lost on me that Google’s AI Overviews handily draw the eye to the most pertinent information, even though I make a conscious effort to scroll past them.
Sam Altman, OpenAI’s controversial CEO, tells us that “AI will continue to get way more capable”; “[p]eople are using it to create amazing things. If we could see what each of us can do 10 or 20 years in the future, it would astonish us today.” Will he one day be similarly credited with Bowie’s soothsaying prowess? Unlikely. When the rock legend cast his prescient judgement on the internet, it was in an era of cautious optimism and sheer naivety. Altman’s comments are far from radical or oracular. Anyone with eyes could compare the non-AI internet of just 10 years ago to the AI-saturated one of today and profess that we have absolutely no idea what another 10 years could produce. If the internet, an infrastructure of connection that stores and links human-made data, is an “alien life form”, then what does that make generative AI? Digital dystopias which imagined man as slave to the machine once seemed catastrophising and unprogressive; these days the news readily tells us the numerical probability of our complete annihilation at the hands of AI.
In the face of this volatility and fear, then, one question remains: why do people engage so willingly with AI? Of course, there are those who boycott it completely for a non-exhaustive list of ethical reasons, and many are trying to limit its use as concerns over security grow. But on the whole, it cannot be denied that the numbers speak in its favour. Why do 73% of British adults use AI in their day-to-day lives? Why do 65% of British workers claim to use AI intentionally in the workplace? Why do 95% of British students admit to using AI in some way to aid their work? And, perhaps most interestingly, why have over a third of British adults used AI chatbots to support their mental health and well-being?
The architects of this technology are nothing if not full of praise for how AI can streamline our workloads by speeding up administrative tasks and data analysis. This, they tell us, will leave us more time for our creative endeavours and ‘higher-level’ problem solving (or if you’re a student, more time for your various extracurriculars…).
Using AI for tasks like these might seem like a no-brainer – in more ways than one; studies have shown that outsourcing our thinking to AI programmes is linked to cognitive decline as we atrophy our creative and critical thinking skills. However, generative AI is far more than a dull but dutiful secretary who lessens our admin load, plodding away with machine-like reliability. Indeed, the very nature of AI means that it is designed to appear anthropomorphic; it has been programmed to ‘learn’, ‘understand’, ‘reason’, and ‘think’ just as a human being might. These anthropomorphic qualities are exacerbated by the human-likeness afforded to AI design: uncanny voices, curated personalities, emotionally-expressive language, and human names are just some of the features many AI large language models (LLM) have been provided with. Claude, much like the fate of Alexa, might find itself dropping rapidly down the baby name popularity list for 2026.
For some, AI use might simply be a dependable way to ensure leaving work at 17:00 on the dot every day. But for many, AI has become a companion and friend, a mentor and guide. It has become a source of comfort and support, something people emotionally depend on and form complex relationships with. It is easy to judge those who rely on AI in this way; it is far less easy, however, to accept that so many of modern life’s conditions have led them directly to this point.
Firstly, despite being more digitally connected than ever before, we are all lonelier than ever, a phenomenon which is especially apparent amongst young people: 16-29 year olds are frequently the most likely generation to report feelings of isolation and suffer from a lack of friends. Human social interactions have been replaced by likes and comments, and by living vicariously through those we engage with online. It is no wonder, then, that people find it quite normal to replicate these quasi-human, parasocial relationships with AI chatbots. AI, like social media, occupies that uncanny liminal space between real and fake, something that can give us the illusion of an authentic interaction without the human sincerity a real one might entail. Ironically, those who engage with AI in this way often aren’t using it for its efficiency, versatility, and precision – they are engaging with it for its anthropomorphic qualities and idiosyncrasies. In short, they are looking for the man in the machine.
In our world of instant gratification, people want immediate satisfaction without the inconvenience and messiness that being human entails. People want emotional comfort without the vulnerability of messaging a friend, they want to be listened to without having to empathise in return, and they want to be reassured without the risk of being rebutted. Creating these quasi-human relationships with AI chatbots gives the illusion of community and support without retreating from the bubble of insularity which the modern world so effortlessly facilitates.
Another thing that AI can perhaps ironically appear to provide is certainty. It seems counterintuitive to associate certainty with technology whose unreliability and (often comical) errors are so frequently highlighted, but the extremely specific, reassuring nature of AI answers can appear like a source of truth in an increasingly uncertain world.
We live in a world of misinformation and distortion, fuelled by political polarisation and extremism. This is coupled with an increasing sense of disenfranchisement and socioeconomic insecurity, which is felt across the political and identity spectrum. People are overwhelmed, hopeless, and pessimistic about the future. AI has, whether latently or explicitly, undoubtedly helped fuel so much of our modern malaise, yet paradoxically it also fashions itself as a cure. When human actors seem in constant conflict with each other and ourselves, AI might appear as a source of truth and authority, a superintelligent being that can provide stability amongst all the chaos. It’s no surprise, then, that 1 in 5 British adults report using AI to keep up with the news. For many, qualities like reliability, authenticity, and sincerity are no longer indicative of human behaviour, but are now better associated with automated programmes.
So, just as Winston loved Big Brother, we now love the AI programmes we were once encouraged to be wary of. Indeed, with their human expressions and constant reassurance, they are so much more familiar than the automated machines we were introduced to in Isaac Asimov’s I, Robot. The unpredictable and pernicious nature of AI has been moulded into the shape of a reliable friend or kindly therapist. It is far more of an “alien life form” than Bowie’s internet, yet it is presented to us in such familiar terms. In a world characterised by uncertainty and instability, it is understandable why so many turn to their chatbots in moments of despair and confusion. Yet, we should remember that our favourite qualities of AI – its apparent capacity for warmth, empathy, and understanding – were copied from a far more trustworthy source: humankind.