The past month has made me realise I’ve been living in a bit of an anti-AI bubble.
That doesn’t mean I haven’t kept up with which edition of Chat-GPT we’re at, like it’s the newest shiny little iPhone we all can’t get our hands off of. I’ve read tons of articles, some arguing against and others for. Ethical implications, labour concerns, environmental impacts — one finds it impossible to be unaware of how this has snowballed into something massive, entirely uncategorisable.
Still, it was easy to naively ignore the extent of AI use. The findings in the media were reminiscent of bad horror stories, discouraging but dismissible as a product of the West. As if their mistakes, their delusions of grandeur, were not adopted by us.
The techbros of Silicon Valley were not at my doorstep just yet. Because they didn’t need to. Not when they were at my bedside, in my pocket, held between my hands.
I open Twitter threads to see more Grok mentions than people. The Straits Times, my local newspaper, gives AI-generated summaries of its articles. Fifteen-year-olds are using ChatGPT to critique their essays before submission.
Forgive me for not understanding its draw. I first used ChatGPT when it entered the mainstream, out of curiosity, and subsequently never opened the website again. To me, it was startlingly clear that AI offered nothing beyond human ability. Of course, there were things I didn’t know, things that could be rectified with an instantaneous response.
But I never felt the need to type that prompt in, because it seemed obvious that in saving time, I would lose out on something else. In asking for targeted research articles, I would miss the chance to go down Wikipedia rabbitholes that never seemed to end. In asking for point-form generalisations, I would reduce someone else’s hard work into mishmashed sentences devoid of any meaning beyond the semantic.
Last week, however, I ended up lying that I had used ChatGPT to study in school, surrounded by my soon-to-be university peers who seemed all too ready to advocate for its use, and I, eager to be a part of this inside joke.
The absurdity of that only hit me much later on the train ride home.
As columnists and professors online criticised the impact of AI on education, my social circle was quickly shifting towards one that embraced it as a supplement to daily life. (This could easily be turned into a critique of the algorithm creating my echo-chamber, but we’ll go into that another day.)
So who’s right and who’s wrong? Who — or what, do we go to for these answers? ChatGPT 5, which promises PhD-level answers on every single topic?
For all my ignorance, I can tell you my answer to the above is a definite no. That any problems affecting real people, no matter how difficult, should never be outsourced to a machine. The moment we begin trusting a chatbot more than we trust ourselves, or the people around us, is when we truly start veering off the path.
Perhaps this is the same anxiety that pervaded the older generation when handphones were introduced. Perhaps I will scoff at nineteen-year-old me ten years down the road, a girl catastrophising something revolutionary.
Still, I cannot help but think there is something severely alarming with mothers encouraging their toddlers to talk to ChatGPT more than they do their family. I do not blame the mothers themselves, many of whom are perhaps struggling with the responsibility of raising a child, working a full-time job, finding time for themselves, and everything in between.
I just worry about what this represents. That something as important as child-rearing, which most would agree benefits greatly from parent-child connection, could be potentially replaced by a chatbot. What does that say about how we view AI?
While the word ‘chatbot’ evidently means tech, it implies a social aspect to the machine, that there is no difference from ‘chatting’ to another individual, as there is conversing with a computerised string of numbers. Except that this ‘friend’ of yours also happens to know everything, and provides 24/7 support.
There is immense danger in this. What used to be either inner dialogue with oneself or conversation with others is now partially diverted to AI, robbing us of the time to self-reflect, to debate, and to consider differences. There is only the eternally correct ‘friend’, the unshakeable truth of lines of code designed to give you the truth, not because it should, but because you want it. Friction engineered out of communication, until we begin to expect it everywhere else.
Why endure the awkwardness of small talk with an acquaintance, or bear the brusque critique of professors, when you have the never-ending sunny optimism of Gemini and the untestable patience of Claude?
For all its sycophantic tendencies and misinformation, one could say AI makes considerably fewer mistakes than a human, considering the sheer volume of communication it deals with. But that doesn’t mean it’s superior, or even anything close to an equal.
Seeing it as both a friend and yet almost God-like in its knowledge makes us all the more unaware of its mistakes, and more believing in its untruths. We won’t realise when it’s replicating the biases of the majority, discriminating against those who don’t speak the language of the Western-centric mold that it is constructed on. When we read the texts we send, the essays we spend hours on, only to realise our voice has been lost, subsumed into the abyss of other people’s stolen work.
To us, it is an assistant, a friend, perhaps even a lover. An ‘intern with the boundless energy of a golden retriever’, as one writer put it. To the companies that run this technology, however, we are just one more user whose attention equals profit, whose cognitive debt incurred from LLMs translates into real dollar bills. We rely on AI to plan our schedules, to ease our worries, crippling ourselves.
Humanity handed over to a machine, starving for more. Yet, it will never truly understand the hunger that we are meant to feel for connection, or a thirst for knowledge. Hunger that can only be sated by genuine experiences, and not prompted into perfection.
There is still much for us to learn about AI, and without it. I just hope it’s more of the latter.