‘It’s free money’: Evergreen student employees use AI to write training dialogues

Student Employees Using AI: Current and former Evergreen student employees have been using AI to generate dialogues for the project’s chatbot, despite Evergreen’s policy prohibiting such practices.
Evergreen Project Overview: Evergreen is a mental health initiative aimed at providing wellness check-ins and Dartmouth-specific social advice through an AI chatbot, set to scale with personalized data.
Policy and Oversight: Evergreen project lead Lisa Marsch asserts that any student using AI-generated content knowingly would be in clear violation of workplace policy and would be immediately addressed.
Detection and Enforcement: The onboarding documents state that students found using AI receive warnings, with multiple offenses leading to job termination, although employees claim the detection isn’t airtight.
Employee Experience: Employees claim that AI-generated dialogues often receive favorable ratings, despite the policy, indicating potential flaws in the review and detection system.
Concerns of Former Advisor: The former Special Advisor to the Provost on AI, James Dobson, highlighted that AI-generated training data might reduce the diversity of responses in the chatbot and its responsiveness to human interactions.

You may have missed