Cornell researchers have found the developers of large language models and the educators who use them have different ideas about the potential harms they may cause, a finding that researchers say underscores the need for educators to be more involved in the tools’ development.
