Just talking about my field specifically, but LLM agents are used in educational institutions to support students with both the course and admin sides of things. They’re doubly handy because they’re also used to collate all of the interactions and identify common blind spots and flag students that are at risk of failing.
I would absolutely not call using a stochastic text prediction machine that has been repeatedly exposed for bias related to gender, age, race, language and topic/context a good thing when applied to students’ grades.
Just talking about my field specifically, but LLM agents are used in educational institutions to support students with both the course and admin sides of things. They’re doubly handy because they’re also used to collate all of the interactions and identify common blind spots and flag students that are at risk of failing.
I would absolutely not call using a stochastic text prediction machine that has been repeatedly exposed for bias related to gender, age, race, language and topic/context a good thing when applied to students’ grades.
It’s not used for marking, obviously.