Against the Quantification of Integrity
When the measure of language becomes its target, it ceases to be good language.
💡Nerd Rating: 1/5. I discuss the origins of certain linguistic tics in LLMs and what it means for writing, student assessment, and thinking.
"It's not x, it's y."
Large Language
this article has a lot of good quotes, but this one really stuck with me:
this is closely related to my pet peeve with academia post The Thing™. im being forced to use AI because you can write more and paradoxically sound less like AI by using AI. the substance of your assignment is completely irrelevant, no one has time to understand what you have written. it has to look right, not actually be right. in fact if you make it more complex, give more nuance or provide more context, you are risking the grader (or more realistically their LLM) not understanding your argument, or worse giving them a wider surface area of attack to stick their pedantic hooks into.
the fact that i used the phrase “surface area of attack” is a disgrace. this relationship shouldn’t be this hostile. i am writing twice as much and learning as half, while getting worse grades.
what is the point of this charade? professors use LLMs to deliberately craft assignments that are supposed to be harder for LLMs but actually punish human effort more, and then use another LLM to guard against LLM use but the ones getting punished are those that didn’t use a LLM to hide the fact they are using a LLM. it’s asinine, that’s the only word i can think of, it’s asinine.
Students should boycott assignments until it stops being like this that’s ridiculous.
This has been a longstanding problem in academia hasn’t it? Papers published (far before AI) just because they looked right and the team paid the magazine the exorbitant fees to get published, but could not be reproduced or made no sense.
peer review always had an issue. im talking about the assignments that professors hand out, which is your primary work for the first 4 to 5 years before doing any amount of actual research. this will also exacerbate the issue papers have when the people who had to start in this environment begin their research career, and in fact it has already gotten worse.