As technology develops and artificial intelligence (AI) becomes increasingly capable, there is a growing potential for AI to be used to cheat on homework and examinations. The maker of ChatGPT – a tool designed to enable users to generate natural language from prompts – has recently announced a new tool to help teachers detect if a student, or AI, wrote the given homework.
The chat-oriented AI of ChatGPT is capable of understanding misspellings and incorrect grammar and can generate answers to given prompts, making it relatively easy for students to pass off AI-generated work as their own. To combat this issue, the new tool will enable teachers to identify if the work has been generated by AI or if it is the student’s own work.
The tool works by analyzing the text submitted and comparing it to the content available in its AI database. It can identify inconsistencies between the submitted text and the content in the database, so if the text submitted doesn’t match up, it is likely the work of a student.
The maker of ChatGPT claims that the tool can accurately identify AI-generated answers with around 90% accuracy. While this isn’t foolproof, it should give teachers an extra layer of reassurance when grading homework.
This new tool should help to reduce the risk of students cheating and improve the fairness of homework and examinations. While not perfect, it is an important step forward in the battle against cheating technology.
Leave a Reply