Dan Kennedy on Nostr: I plugged three paragraphs of a student's paper into ChatGPT 3.5, asked if they had ...
I plugged three paragraphs of a student's paper into ChatGPT 3.5, asked if they had been written by AI, and got a message that they had been written entirely by ChatGPT 3.5. Then I plugged in another three paragraphs and got the same message. As a control, I plugged in three of my own paragraphs and was informed that they were not written by AI. How reliable is this? Can I conclude definitively that the student used AI to write those six paragraphs?
Published at
2023-11-29 16:43:10Event JSON
{
"id": "dd54e3689f874e9e57493b5daa4af4f23fbf9a56d1ec68d9dd61709ebe968fc1",
"pubkey": "c6414669e7665e8bf09b7d8bb3a3e6827f2ca1afaeb1cead0879822a0fbc6a5b",
"created_at": 1701276190,
"kind": 1,
"tags": [
[
"proxy",
"https://journa.host/users/dankennedy_nu/statuses/111494836436266310",
"activitypub"
]
],
"content": "I plugged three paragraphs of a student's paper into ChatGPT 3.5, asked if they had been written by AI, and got a message that they had been written entirely by ChatGPT 3.5. Then I plugged in another three paragraphs and got the same message. As a control, I plugged in three of my own paragraphs and was informed that they were not written by AI. How reliable is this? Can I conclude definitively that the student used AI to write those six paragraphs?",
"sig": "a141af8139a60b0d4a852b85285f9bd16b02ec517d1c8362ee9fb540f3850caff424f6fafb0052090723448ad77f090869dd4d78a4dc1e3b027e937908c7ccf0"
}
