Conclusion
Rehan N -
For the last time, hey everyone! It’s hard to believe that this is the last blog post of my whole senior project journey. Over the past few months I have explored a question that initially seemed kind of niche, but turned out to have tons of real world relevance: how can AI detectors be more equitable for ESL writers.
At the beginning, it was mainly curiosity that guided me as I wondered how AI detectors worked and whether they could be trusted. As I went on, I found during my comparative analysis of both phases of testing, that the bias is very real, and ESL posts had a false positive rate 4-5x higher than native English posts. However, modifying the training data resulted in a significant drop in false positives.
However, this project taught me a lot more than just statistics and coding. It’s taught me about ethical designing in technology being non-negotiable and that being a researcher takes patience even when progress is slow.
I want to give a huge thank you to Ms. Ainslie, my AP research teacher, for guiding me through every twist and turn of this project. I’m also incredibly thankful for my peer reviewers and everyone who commented on my blog. You guys have all shaped something better than I could have imagined.
If you guys would like to hear more about the results of my presentation, feel free to join me for my final presentation on May 13th, 2025 at 2 p.m. Hope to see you there!
Comments:
All viewpoints are welcome but profane, threatening, disrespectful, or harassing comments will not be tolerated and are subject to moderation up to, and including, full deletion.