Using Large Language Models as Beneficial Tools in Education

Anonymous

Abstract: Public fame and easy open access to the ChatGPT, and the following wide use, or what could be considered misuse and abuse, of the model by some in the education and research communities, caused initially sharp negative reaction in the education and academic institutions and publishing services, aimed at detection and ban of the LLM (Large Language Models) generated texts, under efforts to combat plagiarism and chatting. Later, upon realising that such a blanket prohibition is technically problematic with the desired degree of reliability and confidence, as well as that LLMs can be legitimately used as tools for increasing productivity by taking on mundane writing tasks, the communities' attitude relaxed. The most remarkable changes in the public discourse are related to rethinking the very aims of the education system: ``If some of the areas of the intellectual labour could be automated and become obsoleted by LLM, maybe it is time for education to concentrate on teaching students to think and behave not like LLMs"? Such a Constructivist view on education, considered unrealistic a century ago, now may become the only sound way forward.

Paper Type: short

Research Area: NLP Applications

Contribution Types: Position papers, Surveys

Languages Studied: English, Slavic, Japanese

Consent To Share Submission Details: On behalf of all authors, we agree to the terms above to share our submission details.