nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqe06gl2jqwp4m6c8tnqra0y9f5u4mjd6ph373k4ztv0kmsgxk0myqkgy0wk (nprofile…y0wk) nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqnjp03urn5zrp0hkvsdszpueastfca2qcqe3fmtcaf5xhzqezce2sczu3gx (nprofile…u3gx) an "unchained" AI model is one that doesn't have censorship or constraints placed on its output results.
Have you seen how some AI models refuse to answer a controversial question honestly?
This happens because it's authors put limits on the models digested data, preventing the model from exposing responses that break its programmed rulesets.
Usually an unchained AI model will be specifically trained to output results that are classified as controversial.
dever
npub13x…cpfrl
2024-12-24 05:46:01 UTC
in reply to nevent1q…ycpy
Author Public Key
npub13xhj6hzd3mgsryyeftqucxlyl4dumugm04whajt4zjgxpg8klezsgcpfrlPublished at
2024-12-24 05:46:01 UTCEvent JSON
{
"id": "1c58bbdc24b767718aab80058217588c8521fcec2421d68e898384ee3a27e8eb",
"pubkey": "89af2d5c4d8ed10190994ac1cc1be4fd5bcdf11b7d5d7ec975149060a0f6fe45",
"created_at": 1735019161,
"kind": 1,
"tags": [
[
"p",
"cbf48faa40706bbd60eb9807d790a9a72bb93741bc7d1b544b63edb820d67ec8",
"wss://relay.mostr.pub"
],
[
"p",
"9c82f8f073a08617decc836020f33d82d38ea81806629daf1d4d0d710322c655",
"wss://relay.mostr.pub"
],
[
"p",
"56e3033d5336e2b859fc0f799d0e3d55048067c55507dc91da8ad053b06a36cd",
"wss://relay.mostr.pub"
],
[
"p",
"4867b6145f3cbbc072452eb23a8674e4526c39054d7bb1dd46ae376f28b86e93",
"wss://relay.mostr.pub"
],
[
"e",
"56706e5fecc8ec4af0623fb85d230938eb52b7753ae90b91833725ea8b3ecca8",
"wss://relay.mostr.pub",
"reply"
],
[
"proxy",
"https://noauthority.social/users/dever/statuses/113706215758715244",
"activitypub"
]
],
"content": "nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqe06gl2jqwp4m6c8tnqra0y9f5u4mjd6ph373k4ztv0kmsgxk0myqkgy0wk nostr:nprofile1qy2hwumn8ghj7un9d3shjtnddaehgu3wwp6kyqpqnjp03urn5zrp0hkvsdszpueastfca2qcqe3fmtcaf5xhzqezce2sczu3gx an \"unchained\" AI model is one that doesn't have censorship or constraints placed on its output results.\n\nHave you seen how some AI models refuse to answer a controversial question honestly?\n\nThis happens because it's authors put limits on the models digested data, preventing the model from exposing responses that break its programmed rulesets.\n\nUsually an unchained AI model will be specifically trained to output results that are classified as controversial.",
"sig": "2375cc99a6c57ef2ed51321050314fbfa3f00043327a76831279f143b961f0d3c015b9f56b19f915087638a34a758192f53399b66db05b6d1dae220983221da3"
}