Leveraging AI Without Becoming Cognitively Dependent: A Student’s Perspective
In recent years, AI has permeated every aspect of Internet life. Though many of us debate the implications of AI, we increasingly have no choice but to use it. AI is no longer a tool we opt into – it’s the environment we operate in; it’s embedded in search, writing tools, social media, productivity platforms. So now the question is no longer about whether or not you use AI (…you almost certainly are) – it’s about what intentional boundaries and strong foundational thinking skills are necessary to avoid atrophying the very cognitive abilities that make humans valuable.
To have this conversation, it is important to acknowledge that not everyone uses AI in the same way. While total avoidance of AI is becoming increasingly unrealistic, some people are leveraging it to enhance existing skills, while others are relying on it in place of developing those skills at all. That distinction matters more than most conversations around AI currently acknowledge.
Senior leaders are using AI as a multiplier to optimize their processes and encouraging those who come after them to do the same. The difference is that, when they were in college, they developed strong foundations in writing skills, analytical thinking, and domain knowledge, which they are now using AI to supplement. To leaders in the workplace, AI is an efficiency tool.
Students now, on the other hand, should not be assumed to be using AI to supplement foundations they already have. Quite the opposite: many current students are progressively learning less and becoming slower in the long term because they are using AI more. College students are now using it to outsource writing, argument structuring, analytical thinking, and overall problem-solving that are essential both in college and in the workplace. It has been alarming to observe deep conversations, original thought, and the ability to sit with complexity devolve into instant answers, surface-level understanding, and “good enough” thinking. For students in college and the workplace, AI is becoming a substitute for thought.
From a psychological perspective, the phenomenon of some using AI to move faster while others use it so they don’t have to move at all is extremely concerning. When too much cognitive offloading occurs because the brain is left unchallenged, parts of the brain responsible for effortful thinking wither. Struggling through desirable difficulty and unfamiliar concepts creates new synaptic connections, which lead to long-term knowledge retention. When everything becomes easier, nothing forces you to get better. Further, weakening critical thinking skills do not just affect our ability to think independently – they also make AI itself less effective. Poorer thinking leads to poorer prompts, weaker analysis, and less reliable outputs.
I am not saying AI is entirely bad; I use it almost every day to optimize school assignments, summarize long readings, or to help me produce assets for my internship faster. In fact, I am using AI to help me outline and edit this blog right now. What I am saying is that, to mitigate some of the negative effects of using it, we should strategize how to use it to complement our skills rather than replace them. How do we go about this? I propose four ways to use AI without outsourcing our ability to think.
1: Teach AI literacy. Given my previous assertion that AI is no longer optional, I suggest that our best bet for regulating its effects, especially on students, is to teach AI literacy. When I was growing up, people were starting to realize that the use of the Internet was inevitable, but that it should not be used without boundaries, so computer labs and media literacy classes were implemented in elementary school curricula. We learned the difference between primary, secondary, and tertiary sources so we could evaluate where information came from. We analyzed factors such as the author, publication date, and country of origin to assess how bias might affect reliability. We reasoned about when it was appropriate to outsource and cite material, and when it was more effective to originate thoughts. At a time when we were transitioning from rifling through books at libraries to sourcing articles on the Internet, these courses were entirely responsible for my age group’s ability to utilize the Internet as a resource without letting it consume our unique voices. The level of preparation needed around AI is no different to achieving the goal of students seeing value in it as a complement to their work rather than a replacement for their thinking. Early elementary and middle school courses should caution students about AI hallucination – where AI makes up sources and facts – so that students can understand how to vet AI-generated outputs. They should also inform students of the sustainability and critical thinking implications of AI use, so students can make informed decisions about when they want to implement it. Lastly, they should help students understand where AI is going, so that they can learn how to leverage it in school and the workplace to maximize efficiency and edit materials they have produced themselves. Employers will expect both AI fluency and independent thinking (especially if they follow step 2). Teach students that the competitive advantage is no longer simply knowing how to use AI. It is knowing when not to use it. Classes that teach these lessons will solidify AI as a tool for optimizing efficiency without deprioritizing human-led critical thought.
2: Reward critical thinking. This one falls primarily on hiring representatives and teachers. We have most definitely gotten used to a fast-paced world, and the idea of slowing down to make sure we are actually learning rather than simply producing seems unthinkable, especially when it might mean falling behind. Further, the fact that we so often conflate good performance with quickness and efficiency can make it really hard to see the good in slower analytical thinking. But I think a huge part of why students are falling victim to replacing their thinking with AI is that they fear falling behind other interviewees or students if they don’t produce the fastest outputs. Senior leaders should adjust their reward systems to reward entry-level applicants and students for strengths in the analytical processes that underlie their results, rather than for the correctness or speed of the results themselves. Interviewees and students can sound polished but lack depth or struggle without AI assistance. If senior leaders learn to look past outputs and more at inputs – the actual thinking of the person solving the problem – they will catch the students who use AI as a supplement rather than a substitute, and who will be a better asset for them in the long term. In the end, critical thinking is a lot harder to teach than the answer to a specific problem, and our metrics for success should reflect that.
3: Understand AI’s core strengths and don’t be AI-phobic. Despite its very real ethical and cognitive implications, AI can be extremely effective for saving time on tedious tasks, improving accessibility to information, and simplifying complex concepts. Additionally, when used correctly, it can actually supplement creative outputs by helping with ideation and refinement of ideas. Professionals and professors should operate within the reality that AI use is inevitable, increasingly embedded in everyday work, and capable of being genuinely useful. Writing up students for academic citations for using AI in your work while simultaneously using AI to develop curricula and grade work feels contradictory to most students. Additionally, as AI improves and becomes harder to detect, the punishment of sending students to the honor council for using AI becomes less feasible. Rather than being AI-phobic, professors should actively construct curricula to control the ways in which AI is used, such as asking students to cite AI, include the specific prompts they used, and turn on version history for online papers. Constructing systems that regulate AI use is far more realistic than attempting to ban it altogether.
4: Be self-aware and use the AI Balance Rule. And finally, one directed more at students and early professionals than their professors and bosses – be self-aware. You can feel when you are using AI to take over your thinking entirely. You can feel when you are writing without understanding structure, coding without understanding logic, or analyzing without forming real opinions; when you feel this imbalance of AI’s thought to your own, take a step back and fill in the necessary gaps that will allow you to take real intellectual ownership of your work. Here is where I propose the “AI Balance Rule” – a framework for students and early professionals to use to ensure they are using AI correctly and in a balanced way.
- Think first, AI second. Attempt before prompting.
- Use AI to refine, not replace. Prioritize AI as an editing tool rather than as a generation tool.
- Interrogate outputs. Don’t accept – analyze what AI tells you (remember the risk of AI hallucination).
- Practice without it. Build raw skill intentionally.
Using this simple framework, you can ensure that AI extends your thinking rather than stands in for it.
While this article is a start, there are many nuanced conversations to continue having about AI’s effects beyond critical thinking. I hope to expand these thoughts into covering AI and sustainability, AI and the political environment, and AI and protecting your intellectual property more explicitly in future articles.
About the Author
Sophia Fife is a junior at Georgetown University studying Psychology, with minors in Business Studies and Justice & Peace Studies. She is passionate about brand strategy, communications, and consumer behavior, and is currently a Brand Strategy & Research Intern at BrandMirror, where she supports client-facing consulting and internal strategic initiatives. Sophia is especially interested in how values-driven branding and storytelling shape trust, engagement, and long-term impact.
