College Professor Hides Prompt to Catch AI Cheaters, Finds Human Nature Is Pretty Much as We Thought
NVIDIA
A college professor hid a specific prompt in an assignment to detect AI-generated submissions, and the results confirmed that human nature hasn't changed much. The experiment revealed that students who used AI to cheat often failed to notice or follow the hidden instructions, indicating that they were not reading the assignment carefully.
A college professor conducted an experiment to catch students using AI to cheat by embedding a hidden prompt in the assignment instructions. The hidden prompt asked students to include a specific phrase or action, but those who relied on AI to generate their answers often missed it because they didn't read the assignment fully. The results showed that AI-generated responses lacked the personal touch or the specific detail, indicating that the students were not engaging with the material. The professor's findings suggest that while AI can help with work, it can't replace genuine understanding and attention to detail. This experiment highlights the ongoing challenge of detecting AI misuse in education and the importance of designing assignments that require critical thinking.
Source: The Register AI/ML —
original