The author suggests that instead of treating this as an arms race, educators should focus on the purpose of assigning essays: to evaluate students' critical thinking and original arguments. They propose alternative forms of assessment like presentations, podcasts, and reflective writing, and incorporating class-generated content that doesn't exist in ChatGPT's training data. The author warns against punitive measures based on unreliable AI detection software, arguing that it could exacerbate existing inequalities in education. They advocate for working with AI rather than against it, to prepare students for a world where humans coexist with thinking machines.
Key takeaways:
- The rise of AI like ChatGPT has led to concerns about students outsourcing their essay writing to these technologies, prompting universities to adopt AI detection software to identify AI-generated work.
- However, these AI detection tools are not infallible and can lead to false accusations of academic misconduct, particularly for students who use other AI tools for legitimate purposes like grammar and spelling checks.
- Instead of treating the rise of AI as an arms race, academics should focus on the purpose of assignments, which is to evaluate students' critical thinking and original arguments, and consider alternative forms of assessment.
- The panic around AI in academia is symptomatic of a broader crisis in UK universities, with punitive measures based on unreliable detection software potentially exacerbating existing inequalities.