In this #blog-style #advertisement, #OpenAI claims that their LLM is "thinking" and "reasoning". Worse, they actually use "think" and "reason" incidentally, as if that has always been what #LLMs do. We know that fundamentally they can't do that, but the general public likes that story. I'm not surprised by them lying but this blatant obviousness shouts "utter desperation". #ShortSell
https://openai.com/index/learning-to-reason-with-llms/
Dmitri Ravinoff /
npub19k…s723x
2024-09-13 11:05:25
Author Public Key
npub19k8z8gykhdq0uepgdlptys34052tf83grvxpz3rexu0upj2m7slqns723xPublished at
2024-09-13 11:05:25Event JSON
{
"id": "f9a98a9b1408bc8c60fce5cc73ed581dd6b8c1e3a3ff3bb4c9b72c4c880bf992",
"pubkey": "2d8e23a096bb40fe64286fc2b242357d14b49e281b0c114479371fc0c95bf43e",
"created_at": 1726225525,
"kind": 1,
"tags": [
[
"t",
"blog"
],
[
"t",
"advertisement"
],
[
"t",
"openai"
],
[
"t",
"LLMs"
],
[
"t",
"shortsell"
],
[
"proxy",
"https://social.tchncs.de/users/toxomat/statuses/113129916011303215",
"activitypub"
]
],
"content": "In this #blog-style #advertisement, #OpenAI claims that their LLM is \"thinking\" and \"reasoning\". Worse, they actually use \"think\" and \"reason\" incidentally, as if that has always been what #LLMs do. We know that fundamentally they can't do that, but the general public likes that story. I'm not surprised by them lying but this blatant obviousness shouts \"utter desperation\". #ShortSell\nhttps://openai.com/index/learning-to-reason-with-llms/",
"sig": "7876f8f1e49fe53f72c11226074499c20658506f018f4e7f1d8afddebe678c2f79295e5f65f48b92b7457649944f959c9c52bb9cde455a481c3f931199ddcafa"
}