more lettering
This commit is contained in:
parent
19ec01baa3
commit
97b9499d30
|
@ -51,7 +51,11 @@ Now consider, what if out of that two hundred only 10 of them are cheaters? Then
|
|||
|
||||
So the efficacy of a tool like zerogpt is almost entirely dependent on whether you think that cheating is prevelant---but even in the best case scenario you will be punishing students who are not cheating!
|
||||
|
||||
Consider, then,
|
||||
I feel comfortable saying, then, that trying to detect whether students are using LLMs with any kind of automated tool is probably useless.
|
||||
|
||||
So what of cheaters? I don't know. I'm still not entirely convinced that we need to worry about it for most classes. Here's part of my reason: we already know that payments to "paper mills" has gone way down since chatGPT became available. See this article: https://restofworld.org/2023/chatgpt-taking-kenya-ghostwriters-jobs/
|
||||
|
||||
This tells me that the people using chatGPT to cheat were already turning in work that wasn't their own. I think it's kind of a wash and we don't need to upend our teaching and make assessment more inaccessible.
|
||||
|
||||
|
||||
|
||||
|
|
Loading…
Reference in New Issue