Joshua Foust 🎮 on Nostr: Why do so many academics think “we asked GPT” is a valid method? Using a ...
Why do so many academics think “we asked GPT” is a valid method? Using a technological system that produces inconsistent output based on opaque processes is not a method. You can’t do participant observation with a language model and you can’t ask it follow up questions. This is so frustrating.
Published at
2024-11-12 11:47:39Event JSON
{
"id": "8ab87903ba57b6272ae6211d3419d8fbac7dce839b82beedd1582e4a304db833",
"pubkey": "b6609df1c2094fde5270bb674f20b226b1a37903988ddb2f642bde380f6f0076",
"created_at": 1731412059,
"kind": 1,
"tags": [
[
"proxy",
"https://appdot.net/users/joshuafoust/statuses/113469820726021037",
"activitypub"
]
],
"content": "Why do so many academics think “we asked GPT” is a valid method? Using a technological system that produces inconsistent output based on opaque processes is not a method. You can’t do participant observation with a language model and you can’t ask it follow up questions. This is so frustrating.",
"sig": "96a475623c86038080b1f9d328e55759861d46839b1245b2ba71775292ef49d6e683c36ed13023bd2a134eff235a673a1f87516c0fb8a5a0754be121dd3a1cc8"
}
