In his 2023 commencement speech, Chancellor Daniel Diermeier emphasized the importance of “daring [to do] something great that might not succeed.” One might only assume this is what the school administration had in mind when they used Chat-GPT to write an email on the shooting that took place at Michigan State University last year. Vanderbilt’s administration indeed took Chancellor Diermeier’s advice and dared to do something great. However, they certainly did not succeed because they forgot to delete the phrase “Paraphrase from OpenAI’s ChatGPT AI language model, personal communication” at the end of the email.
Yes, some might say that feeling the need to use an AI language model to write a message about a tragic human experience tells a lot about the humanity of the administration. Or, some might even say they would rather read an email written by ChatGPT rather than the administrators. However, according to some students, the administration has not given up on utilizing AI language models for the benefit of the school yet.
The conflict in Gaza sparked lots of student reactions across the campus, and as usual, the administration showed their empathy for the tragedy and their deep understanding of the situation in the most heartfelt way possible: sending out emails. While some students were surprised to not see “Paraphrase from OpenAI’s ChatGPT” at the end of the emails, others wondered, did the administration become smart enough to delete the phrase before sending out the emails?
An anonymous student who did not want their name published because of the fear of getting arrested told VPR, “When I didn’t see the phrase at the end, I almost believed the emails were human written. However, in reading the emails, it was clear to me that a human being couldn’t seriously write that.”
The debate over who wrote the emails is still going on with some students using AI detectors online to find out who the real writer of the emails is. So far, no consistent conclusions have been reached by AI detectors, with some of them finding the emails too inhumane, even for language models like ChatGPT. Outraged by these claims, the Language Model Association Organization (LMAO) held a press conference today, stating, “We put lots of effort into our language models to make sure that they output meaningful messages that make sense. In that sense, being associated with those emails is a big shame for our honorable organization.”
Out of frustration, the Vanderbilt administration wanted to shut down LMAO’s press conference by using its police force and then send another email explaining why they shut it down. The plan was canceled at the last second because one of the administrators got confused and asked, “Well, who’s going to write the email if we can’t use ChatGPT?” According to our reliable sources, administrators got even more frustrated when they found out they couldn’t arrest AI language models.
Some students also realized that some of the emails sent at different times contained the same paragraphs which caused a new wave of student reactions with students wanting to use ChatGPT and copy their previous works in their classes just like their school administration. One of the outraged students told VPR, “I don’t even care about all this, I just want to make administrators send more emails to see if they’re still using ChatGPT.” So far, the administration seems to either be smart enough to not fall into this trap, or they simply could not write an email about this issue without ChatGPT.
Some extremist groups such as Students for Technological Futurism and Understanding (STFU) started to advocate for the total control of ChatGPT for school affairs. One STFU correspondent told VPR, “We conducted a poll with 30 students and 10 administrators, and we found that 92 percent of our respondents would prefer if the school was run by ChatGPT. We want change and we want it now.” While the support for groups like STFU continues to grow, Vanderbilt’s future remains uncertain.
Image by Arda Daren Aslan
