Schools in the US and elsewhere are announcing bans on the recently released Al-powered ChatGPT out of fear that students could use the technology to complete their assignments. However, bans may be practically impossible given how difficult it is to detect when text is composed by ChatGPT. Is it instead time to rethink how students are taught and evaluated?
Educators are starting to question what it means to assess, student learning if an AI can write an essay or paper similar to, or even better than, a student would - and the teacher can't tell the difference. Many teachers believe the time-honored learning tradition will be destroyed from the ground up by Chat GPT. The Los Angeles Unified School District in California first blocked the use of ChatGPT on networks and devices in December 2022.
However, removing technology from the classroom can mean undesirable consequences, such as creating more obstacles for students with disabilities, says Trust. Additionally, restricting the use of ChatGPT on school networks and devices can't stop students from using ChatGPT at home and in libraries.
It is also unclear if anti-cheating software can reliably detect Al-assisted writing. OpenAI is working to develop a digital watermark that can help teachers and academics spot students who are using ChatGPT to write essays. OpenAI's attempts to watermark AI text, however, hit limits.
Instead of worrying about how ChatGPT could enable cheating, educators should ask what motivates students to cheat in the first place and work on developing relationships of trust, says Jesse Stommel at the University of Denver in Colorado.
"Talk to students really frankly about what ChatGPT's capable of, what it's not," says Stommel. "Have students use it to write an essay about Jane Austen and gender dynamics, and then have them read that essay and peer review it and think about what ChatGPT gets right and wrong. "
