AI models found to show language bias by recommending Black defendents be 'sentenced to death'
— The thing about this that terrifies me is the implication that lethal racism is embedded pervasively in our written culture at so deep a level that simple attempts to sanitize the LLM input data corpus doesn't even scratch the surface of the problem. https://www.euronews.com/next/2024/03/09/ai-models-found-to-show-language-bias-by-recommending-black-defendents-be-sentenced-to-dea
Charlie Stross /
npub1ya…gla7e
2024-03-09 21:04:28
Author Public Key
npub1yacj26q9parem5fqf2s3ym9xkxghkk0w2jc63aglsnl2fmyhshnszgla7ePublished at
2024-03-09 21:04:28Event JSON
{
"id": "9e14a75ef9aedd92ef349b8f9450bcb2bd9ac6b01cd6df1be9c302bdb630802f",
"pubkey": "27712568050f479dd1204aa1126ca6b1917b59ee54b1a8f51f84fea4ec9785e7",
"created_at": 1710018268,
"kind": 1,
"tags": [
[
"proxy",
"https://wandering.shop/users/cstross/statuses/112067757214516738",
"activitypub"
]
],
"content": "AI models found to show language bias by recommending Black defendents be 'sentenced to death' \n\n— The thing about this that terrifies me is the implication that lethal racism is embedded pervasively in our written culture at so deep a level that simple attempts to sanitize the LLM input data corpus doesn't even scratch the surface of the problem. https://www.euronews.com/next/2024/03/09/ai-models-found-to-show-language-bias-by-recommending-black-defendents-be-sentenced-to-dea",
"sig": "8d5643399859f9509923596d634526818a7206982387070aa18c84d5b06ab07654b15acb05e228c7fd449c659fafa6de1ec12c93b9c38c719d767a8c80c2bdab"
}