Koen Hufkens, PhD on Nostr: Some observations of using LLMs for text based tasks. 1. it mostly fails, at first 2. ...
Some observations of using LLMs for text based tasks.
1. it mostly fails, at first
2. if it finally works, through "prompt engineering", you sound like a micro-managing prick
You got to wonder how LLM use will impact real world social interactions.
#AI #LLM
Published at
2024-10-07 12:27:13Event JSON
{
"id": "c8085b1cfa2f9f93c029827add9ea909b814108ea2b03222cb94241be32af140",
"pubkey": "6638e12cc97dd416403d7ee85307be522399318076ebd2aceba1a4f0cadeab51",
"created_at": 1728304033,
"kind": 1,
"tags": [
[
"t",
"ai"
],
[
"t",
"llm"
],
[
"proxy",
"https://mastodon.social/users/koen_hufkens/statuses/113266133127711799",
"activitypub"
]
],
"content": "Some observations of using LLMs for text based tasks.\n\n1. it mostly fails, at first\n2. if it finally works, through \"prompt engineering\", you sound like a micro-managing prick\n\nYou got to wonder how LLM use will impact real world social interactions.\n\n#AI #LLM",
"sig": "61fa34c62986dd9dbd701cd392fad229aae85e1b0cde7efa46d381668d2ab2e93e7701d4f98e394194b11019430570428fbc39d49bbbb05a65eefbe8e70588ac"
}
