noumenon@lemmy.world to Artificial Intelligence@lemmy.worldEnglish · 3 months agoBBC tested ChatGPT, Copilot, Gemini, and Perplexity on 100 of its own articles; 51% of answers had significant issueswww.bbc.co.ukexternal-linkmessage-square3fedilinkarrow-up123arrow-down10cross-posted to: hackernews@lemmy.bestiver.se
arrow-up123arrow-down1external-linkBBC tested ChatGPT, Copilot, Gemini, and Perplexity on 100 of its own articles; 51% of answers had significant issueswww.bbc.co.uknoumenon@lemmy.world to Artificial Intelligence@lemmy.worldEnglish · 3 months agomessage-square3fedilinkcross-posted to: hackernews@lemmy.bestiver.se
minus-squareAlexstarfire@lemmy.worldlinkfedilinkEnglisharrow-up10·3 months agoThey also say that 91% of answers contained at least some issues. That’s pretty bad.
minus-squareThorry@feddit.orglinkfedilinkEnglisharrow-up1·3 months agoThe other 9% it probably just errored out and didn’t give an answer at all
They also say that 91% of answers contained at least some issues. That’s pretty bad.
The other 9% it probably just errored out and didn’t give an answer at all