Challenges and Prospects for Linguistics in the Context of ChatGPT
The success of ChatGPT in language comprehension and generation challenges some traditional concepts in linguistics.This article addresses three unavoidable issues for linguists in the era of artificial intelligence:(1)how to treat the language generated by large language models such as ChatGPT?(2)how to provide semantic support for artificial intelligence systems such as language models?(3)how to form human-machine collaboration with artificial intelligence systems such as large language models?This article concludes with three tentative propositions:(1)"machine language"is a kind of"interlanguage";(2)denotational semantics and distributional semantics should be combined;and(3)there are three modes of human-machine collaboration,namely,embedding,copilot and agent.
ChatGPTlarge language modelsartificial intelligencesemanticshuman-machine collaboration
万方数据
维普
