"Scaling up and shaping up large language models increased their tendency to provide sensible yet incorrect answers at difficulty levels humans cannot supervise, highlighting the need for a shift in AI design towards reliability, according to a paper in Nature."
https://www.nature.com/articles/s41586-024-07930-y
Something something confirmation bias
Neo-Rodneyite: Yap Szn ✍🏿📖
npub17z…rj73v
2024-09-30 15:44:00 UTC
Author Public Key
npub17z32qtcm93sfaujd7gxh7qrf8rh488yz5mme3andl7ca9zs2uuvsarj73vSeen on
wss://relay.momostr.pinkPublished at
2024-09-30 15:44:00 UTCEvent JSON
{
"id": "a11f4a672b6e1169742766dd1c427cc44558cc0facd61a3a9c6c5aef2e6f4dc2",
"pubkey": "f0a2a02f1b2c609ef24df20d7f006938ef539c82a6f798f66dffb1d28a0ae719",
"created_at": 1727711040,
"kind": 1,
"tags": [
[
"content-warning"
],
[
"proxy",
"https://todon.eu/@jalcine/113227270768291251",
"web"
],
[
"proxy",
"https://todon.eu/users/jalcine/statuses/113227270768291251",
"activitypub"
],
[
"L",
"pink.momostr"
],
[
"l",
"pink.momostr.activitypub:https://todon.eu/users/jalcine/statuses/113227270768291251",
"pink.momostr"
],
[
"-"
]
],
"content": "\"Scaling up and shaping up large language models increased their tendency to provide sensible yet incorrect answers at difficulty levels humans cannot supervise, highlighting the need for a shift in AI design towards reliability, according to a paper in Nature.\"\n\nhttps://www.nature.com/articles/s41586-024-07930-y\n\nSomething something confirmation bias",
"sig": "1af131849aa179ac92720d8ab664e9e562f4fc76501e6240ea28f13ed9aa79ccd9ca717a5a45d474ed15ace89c3db7632966349215a3c00c9fa66967ccfa386d"
}