The Algorithm Isn’t Thinking- You Are
Sydney25 May,2025
OPED BY Biranchi Poudyal PhD Candidate, Charles Darwin University
Game theory fundamentally examines the decisions made by rational actors seeking to maximize their outcomes. And while it is often applied to economic or political science, I believe it is the perfect lens for understanding something closer to my current work: how students engage with generative AI in academic life. As a PhD candidate researching generative AI and its implications in education, I have observed students shifting their relationship with these new learning tools.
The emergence of GenAI models, such as ChatGPT, has sparked a philosophical debate about the implications for education. Many critics fear that students will stop thinking, while others worry about the loss of academic integrity. However, I’d like to offer a different perspective: what if students are not outsourcing their intelligence but simply strategizing more efficiently in a game where the only real opponent is themselves?
Let me explain this with a hypothetical scenario of a student who is preparing to write a research paper. In the traditional approach, they might engage in a multi-step process: visit the library or surf online (Step 1), locate relevant books or online journals (Step 2), read them thoroughly (Step 3), and ultimately begin to understand the material (Step 4). Only after this stage will they start to synthesize, reflect and then create new content. It’s a noble process, but also a slow one, particularly in the age of digital technologies. Here comes the GenAI in a picture. It does not skip understanding; rather, it helps you to start from that point. With the right prompt, a student can access summaries, opposing arguments and case studies in minutes. This is not cheating; it is a strategic entry. The knowledge is not handed over like a completed assignment; it’s unlocked and restructured in real-time, based on the student’s input.
And the most important thing to note is that what the AI gives you is what you provided, just altered through a neural mirror. If you ask shallow questions, you will receive surface answers. If you engage deeply, GenAI will meet you there. So, this process is not merely outsourcing thought; it is scaling it. In game theory terms, the AI is not a competing player because it has neither any goals, nor a strategy, nor a reward to gain. The only rational actor is the student. They are not playing against AI, but against their own impulses, goals and limitations: procrastination, perfectionism, fear of failure, and time pressure. In this scenario, the AI is not a shortcut. It is a tool to outgrow constraints with more clarity and speed.
It is here that the role of AI becomes most interesting as a co-creator. When a student reflects on a topic and prompts an AI model, they are already engaging in the metacognitive act of organizing their thoughts and exploring new ideas. GenAI, in turn, offers a response that will be the foundation for further thinking. This is a self-driven learning process that blurs the traditional boundaries of authorship. There is no denying that the output is generated, but the inputs? It is original, shaped by what the student knows, what they seek to express and how they are framing their ideas. From this perspective, students are using a tool that allows them to see their own thinking reflected back in an organized and clear manner.
Of course, not all uses are ethical or productive, but the same can be said for books. If a student reads only pro-Marxist or anti-Marxist texts, their output will reflect that. The medium, whether a book or AI is not the problem; critical engagement is. And let us be honest here: we have always used tools to think. Even the Socratic method was a tool; the printing press and calculator were revolutionary ones. Google did not replace knowledge; it expanded our creative reach. Likewise, generative AI is no different. Every decision a student made about when, how, and to what extent they would use AI was a decision about their intellectual boundaries. The student’s challenge is not to outsmart the AI, but to move within a matrix shaped by their own decisions. They must choose between speed and depth, between ease and effort, between an answer and an understanding.
So, the real challenge here is not to suppress its use, but to educate its use effectively. We need to teach students how to ask better questions, how to cross-examine GenAI outputs, and how to transition from generation to reflection. We must design assignments that reward creativity, not merely completion. Most importantly, we should stop treating generative AI like a villain and start recognizing it for what it is: a medium that helps students to explore their creative horizons.
Biranchi Poudyal
PhD Candidate, Charles Darwin University
