At what point are our thoughts no longer our own?

By Eliza Bellones

A mere five years ago, the capabilities of publicly accessible AI were largely confined to rudimentary text completions and simplistic artistic outputs. Being somewhat detached from the computing sphere, I considered genuinely intuitive artificial intelligence to be eons away. And yet, I now write this article in a society where many columns like mine are penned not by journalists, but by large language models. This is not to say I find generative AI to be an affliction; its capacity for automation and productivity is unparalleled. The problem arises when AI is used not for mechanization but as a stand-in for human perspective. I have come to ponder the question: at what point does it become too much?

We risk losing the friction of human thought in exchange for convenience. Living through the rise of misinformation and social media in the Philippines has taught me a lasting lesson: when we delegate our thinking, we relinquish our agency. If social media proved how easily our views can be manipulated, the rise of AI reveals how easily they can be replaced. Generative AI invites us to outsource creation, and when we do, it ceases to be a tool and becomes a proxy for the mind. The real threat is not just that we begin to rely on external thought. It is that we may forget to think for ourselves.

This risk is compounded by the illusion that algorithms are neutral thinkers. This is false on both counts: generative AI does not possess a free mind. It is trained on datasets whose biases it inevitably absorbs and mirrors, favoring consensus and sanding off outliers. Free human thought, on the other hand, depends precisely on those outliers. Eccentricity, the ability to challenge the consensus and generate genuinely novel ideas, is the catalyst of human progress—ironically, it may be what led to the development of AI in the first place.

Deep thinking requires wrestling with uncomfortable thoughts until you reach a balanced, empirical conclusion. AI eliminates that by immediately offering a polished answer. While an output may originate from your prompt, knowing what you think is inseparable from knowing how you arrived there. When you ask AI to write or reason for you, you aren’t just letting it choose words; you are letting it choose the premise. And when you don’t come to an outcome through independent thinking, you may not recognize when an AI response subtly nudges your worldview towards its inherent bias.

Let me be clear once again: this is not a scathing indictment against AI. This is simply a reminder to draw the line between using a tool and surrendering our voice. As a writer, I have always held the opinion that words are power. In times like these, the most important thing we can do is to protect our capacity to think for ourselves, because to lose our words is to lose our perspective. Rather than quietly yielding to the echo chamber, we must refuse to outsource the one thing that makes us human: our right to decide what we believe.