top of page

ChatGPT Tried the Ivy League's Essays - Could It Get In?

I loved reading about the NYTimes experiment with ChatGPT trying to pass Ivy League essay prompts. The authors came to the conclusion that AI is not there yet - duh. But frankly, as someone who slogged through those applications myself, I realized my writing was equally formulaic as what those bots produced. And I bet AI will soon be indistinguishable from even exceptional applicants.

Rather than whether AI can game the system, maybe we should question if these essays really reveal meaningful insights beyond grades and test scores. Except for those with incredible stories, most students write painfully generic essays.

In the experiment, ChatGPT generated an okay paragraph about relating to a song for Princeton. But other bots totally faked knowledge of lyrics - showing the risks of AI's tendency to make stuff up.For Yale and Harvard, the bots just regurgitated prompts or gave pleasant but standard responses. The Dartmouth essay relied on cringey admissions clichés.

The takeaway was today's AI still needs improvement to nail essays. But my memory is I also just formatted my writing to sound "smart" in the way I thought colleges wanted. I didn't gain much from the tortured process or think it captured my essence.

I'm positive that in a few years, AI skills will advance enough to flawlessly produce such artificial essays indistinguishable from even the most impressive human applicants. But are those skills what we should define merit by?

Maybe instead of essays, admissions could require a simple video intro or alumni interviews? Direct interactions like those could give a truer picture of applicants beyond performative writing that mostly demonstrates how well they pander.

Of course, video applications risk their own equity issues. Costs of filming professionally could disadvantage some. But recognizing the emptiness of current essays is a start. We can then brainstorm the next evolution - focused more on assessing promise than meaningless box-checking.

My perspective comes from finding the essay process largely hollow. I won't mourn robots mimicking such artificial hoop-jumping. This is a chance to reimagine evaluations. Our potential is more than bot-beatable essays.

Key Takeaways:

  • Essay Authenticity Debated: AI's formulaic essays shed light on similar patterns in human responses, questioning their admissions value.

  • AI's Potential: Current AI limitations may soon evolve to match top human essayists.

  • Time for Admissions Change: Alternatives like video intros could provide deeper insights, despite inherent challenges.

Cover image crafted using Midjourney. Want to see how it was made? Check out the creative prompt used: "Image of a student entering harvard, ivy league university. the student holds a hand of a friendly humanoid looking robot. Hyperdetailed photography, photorealistic. Canon EF 16-35mm f/2.8L III USM lens on a Canon EOS 5D Mark IV camera.****”

Disclaimer: This blog post was authored by a human, but research and editing assistance was provided by artificial intelligence.

60 views0 comments


bottom of page