Keith Wilson on Nostr: AI can easily be trained to lie, and it can’t be reversed using current AI safety ...
Published at
2024-01-15 12:55:19Event JSON
{
"id": "2694b326735987ede28329c319ff64d1f6f6a4dd24eb69a69e0afe78b9cff6fb",
"pubkey": "392ae419f68f523e9bfb15de0d4d774dd6a196ab349a5cbf852f74b94bfcf0fb",
"created_at": 1705323319,
"kind": 1,
"tags": [
[
"t",
"aiethics"
],
[
"t",
"aisafety"
],
[
"proxy",
"https://fediphilosophy.org/users/keithwilson/statuses/111760069065302226",
"activitypub"
]
],
"content": "AI can easily be trained to lie, and it can’t be reversed using current AI safety measures. #AIEthics #AISafety https://apple.news/AJpHohFEdRw-YSGRJXkDfyw",
"sig": "030bf9dfde16052cf6533d001346fcd9da01dee2d7ebb48f71963de1746c45ce51bbdf41b41738769d3ae667d4bb7beb3419bd99e6418d8018675963d18a8eda"
}
