Reconsidering AI’s Place at SIPA
By Stephen Chmelewski
During the first few minutes of a SIPA class last year, our normally lighthearted professor became very serious; several students in the class had written nearly identical reflection papers on the week’s readings. The professor speculated that this could have been the result of students working together, or, more likely, that AI had devised similar analyses for them of the week’s texts.
One could hear a pin drop after these accusations were leveled, and yet the room somehow became even quieter when our instructor said, “If all of you are unable, or unwilling, to write a one-page reflection paper in graduate school without the use of artificial intelligence, then I invite you to reflect on what it is that you’re doing at this institution.” Damn.
I don’t know if my peers relied on AI, and nor am I interested in reprimanding them for it, but I have been grappling with the professor’s remarks since that lecture. If ChatGPT, Claude, and Gemini are coming up with arguments for us, writing our essays, and generating in-class talking points on our behalf, then what exactly are we doing at SIPA? Why immerse ourselves in this academic environment and spend thousands on tuition if we’re going to treat prompts for learning as inputs for large language models?
There is no “correct” reason to be at SIPA, and each of us may, in fact, have several motives for studying in IAB. Perhaps, like me, you’re here to meet interesting people, build professional connections, and also to develop new ways of thinking and communicating about complex global problems.
If this resonates—if you’re also at SIPA to strengthen your critical and analytical thinking—I encourage you to consider how your relationship to AI may be impacting your ability to achieve that goal.
When we consistently use AI to generate new ideas or to write (or speak) for us, our capacity for thinking and ideating can take a massive hit. Our analytical reasoning becomes almost entirely sidelined when ChatGPT is asked to produce “our” thesis statements, string together “our” paragraphs, and come up with “our” talking points for a seminar discussion. I use quotations because what’s generated by AI isn’t ours; rather than coming from our synapses, these arguments and these words bubble up from the depths of a large language model.
A 2025 study by MIT Media Lab found that, “while LLMs offer immediate convenience, (...) users consistently underperformed at neural, linguistic, and behavioral levels” and that this “(...) raise[s] concerns about the long-term educational implications of LLM reliance and underscore[s] the need for deeper inquiry into AI's role in learning.”
Despite these findings, these technologies can actually bolster and expand our learning if leveraged strategically. For example, AI can be used to uncover new sources for a research topic (just be sure to cross-check them), or to translate scientific jargon into more easily digestible language. When used in these ways, AI can surface new information and broaden one’s knowledge base, adding a valuable tool to the graduate student’s educational toolbelt.
The ubiquity of AI use at SIPA is unsurprising given its usefulness. Having it think and reason for us is undoubtedly easier than completing classwork ourselves, and it certainly frees up time to do the things we’d rather be doing. (Don’t tempt me, Arts and Crafts). Maybe it can even help boost our scores on a few assignments. With all of its perks, AI is a deeply alluring technology for busy, stressed-out, exhausted policy students like ourselves.
Given the hecticness of SIPA life, I certainly could have drafted this article with AI and saved myself a significant amount of time, and yet I’m glad that I didn’t. Writing this was challenging and time-consuming, but that’s because it forced me to think meaningfully about a topic I care about and to find ways to articulate my thoughts. Plus, I feel like I’ve really earned that trip to Arts and Crafts now.
Whether you’re at a bar with friends or simply walking across campus alone, take a moment at the dawn of this academic year to “reflect on what it is that you’re doing at this institution.” Take stock of the ways in which AI could complement your academic journey and your intellectual ambitions without suppressing your creativity or dimming your critical thinking. Maybe that means utilizing Claude strictly in the fact-finding stages of a project, or letting Gemini develop your weekly study schedule. Perhaps that means refraining from AI use entirely during your time at SIPA. That’s your balance to find.