Rather than a single rule that is impossible to keep, Cheshire Academy codes its guidelines like traffic lights: free AI, partially authorized AI, or a complete ban.
Between organizations that encourage the adoption of AI in the classroom, from OpenAI to UNESCO, and teachers who still don't know what to allow, the gap remains wide. Cheshire Academy, a Connecticut school that serves about 400 students in grades 9 to 12, has chosen a deliberately simple answer: each assignment has a color. Green means AI is fully allowed, red means it's off-limits, and orange opens up an in-between where the teacher allows certain tools — a spell checker, for example — while discarding chatbots.
The approach is accompanied by three principles in the field. The first consists of taking students where they are, admitting that experimentation will take place anyway and preferring education to disinformation. The second is exemplary: to show in practice that AI excels at automating tasks but struggles with the precision and the author's own voice. The third requires systematically checking the quotes and statements produced, as hallucinations can lead a learner astray for a long time.
The school has also set up a student council dedicated to AI, responsible for leading the discussion on healthy use among peers, and trains its teachers in the formulation of requests as well as the limits of technology. In terms of tools, the school relies on the MagicSchool platform, whose individual formulas remain under the $100 per year mark.
Read the full article on the original source
Go to the source →