"GPT-3, '공격적' 언어를 선호하는 인간을 모방한다"

Chatbots are getting better at mimicking human speech, but are also getting better and for worse. A new study of Reddit conversations found chatbots replicate our fondness for toxic language, with two prominent dialogue models twice as likely to agree with 'offensive' comments than 'safe' ones.

GPT-3 mimics human love for ‘offensive’ Reddit comments, study finds


기자의 다른 기사보기
저작권자 © Tech42 - Tech Journalism by AI 테크42 무단전재 및 재배포 금지

관련 기사

인공지능의 차세대 개척지 '영화·TV 더빙'

A new class of start-ups is hoping to address the unintentional performance art of dubbed Hollywood movies. Using artificial intelligence...

아마존 슬라임, 제품 도용에 '검색 결과 조작'

Amazon ripped off its third-party sellers’ products and rigged search results to favor its own products even as Jeff Bezos...

구글, 기후변화 부정 콘텐츠 '홍보·광고' 금지

Google is cracking down on digital ads promoting false climate change claims or being used to make money from such...

Microsoft, 중국에서 '링크드인 철수'

Microsoft Corp.’s LinkedIn will shut the version of its professional-networking site that operates in China, marking the last major American...