Matches in SemOpenAlex for { <https://semopenalex.org/work/W4385852384> ?p ?o ?g. }
Showing items 1 to 93 of
93
with 100 items per page.
- W4385852384 abstract "Abstract We investigate the political bias of a large language model (LLM), ChatGPT, which has become popular for retrieving factual information and generating content. Although ChatGPT assures that it is impartial, the literature suggests that LLMs exhibit bias involving race, gender, religion, and political orientation. Political bias in LLMs can have adverse political and electoral consequences similar to bias from traditional and social media. Moreover, political bias can be harder to detect and eradicate than gender or racial bias. We propose a novel empirical design to infer whether ChatGPT has political biases by requesting it to impersonate someone from a given side of the political spectrum and comparing these answers with its default. We also propose dose-response, placebo, and profession-politics alignment robustness tests. To reduce concerns about the randomness of the generated text, we collect answers to the same questions 100 times, with question order randomized on each round. We find robust evidence that ChatGPT presents a significant and systematic political bias toward the Democrats in the US, Lula in Brazil, and the Labour Party in the UK. These results translate into real concerns that ChatGPT, and LLMs in general, can extend or even amplify the existing challenges involving political processes posed by the Internet and social media. Our findings have important implications for policymakers, media, politics, and academia stakeholders." @default.
- W4385852384 created "2023-08-17" @default.
- W4385852384 creator A5016703677 @default.
- W4385852384 creator A5018012312 @default.
- W4385852384 creator A5080836354 @default.
- W4385852384 date "2023-08-17" @default.
- W4385852384 modified "2023-10-02" @default.
- W4385852384 title "More human than human: measuring ChatGPT political bias" @default.
- W4385852384 cites W1494689044 @default.
- W4385852384 cites W1935981421 @default.
- W4385852384 cites W2014229506 @default.
- W4385852384 cites W2014580705 @default.
- W4385852384 cites W2060704337 @default.
- W4385852384 cites W2117442186 @default.
- W4385852384 cites W2171773113 @default.
- W4385852384 cites W2770851335 @default.
- W4385852384 cites W2782907140 @default.
- W4385852384 cites W2800290766 @default.
- W4385852384 cites W2811198947 @default.
- W4385852384 cites W2891525068 @default.
- W4385852384 cites W2893425640 @default.
- W4385852384 cites W2896508999 @default.
- W4385852384 cites W3093268905 @default.
- W4385852384 cites W3122056237 @default.
- W4385852384 cites W3123342105 @default.
- W4385852384 cites W3123948795 @default.
- W4385852384 cites W3134323617 @default.
- W4385852384 cites W3195992089 @default.
- W4385852384 cites W3196248941 @default.
- W4385852384 cites W3199135928 @default.
- W4385852384 cites W4206590911 @default.
- W4385852384 cites W4221031488 @default.
- W4385852384 cites W4285129823 @default.
- W4385852384 cites W4287887133 @default.
- W4385852384 cites W4287887510 @default.
- W4385852384 cites W4297154096 @default.
- W4385852384 cites W4298125028 @default.
- W4385852384 cites W4319083882 @default.
- W4385852384 cites W4319167005 @default.
- W4385852384 cites W4361004969 @default.
- W4385852384 cites W4361855702 @default.
- W4385852384 cites W4363624465 @default.
- W4385852384 cites W4376643691 @default.
- W4385852384 cites W4385364884 @default.
- W4385852384 doi "https://doi.org/10.1007/s11127-023-01097-2" @default.
- W4385852384 hasPublicationYear "2023" @default.
- W4385852384 type Work @default.
- W4385852384 citedByCount "3" @default.
- W4385852384 countsByYear W43858523842023 @default.
- W4385852384 crossrefType "journal-article" @default.
- W4385852384 hasAuthorship W4385852384A5016703677 @default.
- W4385852384 hasAuthorship W4385852384A5018012312 @default.
- W4385852384 hasAuthorship W4385852384A5080836354 @default.
- W4385852384 hasBestOaLocation W43858523841 @default.
- W4385852384 hasConcept C118084267 @default.
- W4385852384 hasConcept C138921699 @default.
- W4385852384 hasConcept C15744967 @default.
- W4385852384 hasConcept C162324750 @default.
- W4385852384 hasConcept C17744445 @default.
- W4385852384 hasConcept C193521449 @default.
- W4385852384 hasConcept C199539241 @default.
- W4385852384 hasConcept C39549134 @default.
- W4385852384 hasConcept C77805123 @default.
- W4385852384 hasConcept C90508534 @default.
- W4385852384 hasConcept C94625758 @default.
- W4385852384 hasConceptScore W4385852384C118084267 @default.
- W4385852384 hasConceptScore W4385852384C138921699 @default.
- W4385852384 hasConceptScore W4385852384C15744967 @default.
- W4385852384 hasConceptScore W4385852384C162324750 @default.
- W4385852384 hasConceptScore W4385852384C17744445 @default.
- W4385852384 hasConceptScore W4385852384C193521449 @default.
- W4385852384 hasConceptScore W4385852384C199539241 @default.
- W4385852384 hasConceptScore W4385852384C39549134 @default.
- W4385852384 hasConceptScore W4385852384C77805123 @default.
- W4385852384 hasConceptScore W4385852384C90508534 @default.
- W4385852384 hasConceptScore W4385852384C94625758 @default.
- W4385852384 hasFunder F4320321091 @default.
- W4385852384 hasLocation W43858523841 @default.
- W4385852384 hasOpenAccess W4385852384 @default.
- W4385852384 hasPrimaryLocation W43858523841 @default.
- W4385852384 hasRelatedWork W2132175522 @default.
- W4385852384 hasRelatedWork W2343746689 @default.
- W4385852384 hasRelatedWork W2364314632 @default.
- W4385852384 hasRelatedWork W2366819613 @default.
- W4385852384 hasRelatedWork W2748952813 @default.
- W4385852384 hasRelatedWork W2788069684 @default.
- W4385852384 hasRelatedWork W2899084033 @default.
- W4385852384 hasRelatedWork W3114741055 @default.
- W4385852384 hasRelatedWork W4280562317 @default.
- W4385852384 hasRelatedWork W4382728419 @default.
- W4385852384 isParatext "false" @default.
- W4385852384 isRetracted "false" @default.
- W4385852384 workType "article" @default.