Some faculty members enthusiastically embrace a future in which these programs become part of everyday life, because they can help reduce time on routine tasks, function as a personal tutor, or kick start ideas for essays and research papers.
Many — including some of these AI enthusiasts — are deeply worried that students have been handed a powerful tool that comes without a training manual, so there’s no way to tell whether it’s producing insights or inaccuracies. They also fear that students may become willing to cede the difficult work of critical thinking and analysis in favor of a time-saving device that can churn out a quick discussion post or essay.
These visions of the future run the gamut. But we wanted to know where things stand now. Have classrooms been overrun with AI-produced work, or has ChatGPT been a useful tool? Do students understand what generative AI is and how to use it properly? Are instructors changing how they teach?
We asked. You answered. More than 70 people wrote in to describe what they experienced the first semester in a world where these tools have become easily accessible.
If you haven’t read our story yet, we encourage you to do so. Whether you’re excited or worried or somewhere in between, you’ll probably find your views reflected in some of what you read. Here are a few takeaways:
Obvious cheating with AI was present but not pervasive. For the most part, professors said they could spot the work that had been AI-generated because, say, the tone was radically different from students’ earlier writing, it included material not discussed in class, or it was flat-out wrong in describing or summarizing concepts and content.
Of course, readers also noted that there are ways in which students could have reworked AI-generated text to make it less detectable. That’s only likely to be even more true as students become familiar with these tools.
Figuring out whether students definitely cheated with AI was time consuming. While AI writing can be easy to spot, faculty members needed to be sure of it if they were going to confront students. Given that nobody has invented a foolproof AI detector, professors said they had to spend a lot of time studying students’ writing and looking for other clues. Many used Turnitin’s AI detector or some similar application, but treated it as a starting point for further exploration rather than the final word.
Most professors plan to change their teaching next semester. One obvious place to start is to have a clear AI-usage policy. Almost 80 percent of respondents plan to do that. This in itself will be tricky to navigate because some instructors don’t want students to use any AI and others are OK with responsible — and transparent — use. About 70 percent also plan to change their assignments to make it harder to cheat with AI, such as doing more work in class.
Guidance is lacking. Readers described having to navigate these challenges on their own. They are looking for support in figuring out where the line is between appropriate use and cheating, altering their course design, and helping students understand what these tools are and what they can and can’t do well.
Thank you to everyone who wrote in and shared your thoughts and experiences. We appreciate it!
Your plans for AI
Colleges have yet to come to terms with how generative AI will affect undergraduate teaching and learning. That’s not surprising, given how recently ChatGPT appeared on the scene. But this summer will be a critical period for many to rethink course design, refashion academic-integrity policies, and create or revise courses to respond to how AI will reshape a varied range of disciplines, including computer science and the arts.
We want to hear from you. What is happening in your department, your school, or on your campus when it comes to AI? We’re particularly interested in hearing from places that are taking a holistic approach to these discussions. Are you creating new coursework to help students understand a future with AI? Is your department working together to redesign introductory or writing-intensive courses? If you think you’ve got something particularly innovative or forward-looking happening, write to me at beth.mcmurtrie@chronicle.com.
You can also fill out our Google form, which we’ve updated with a question on what kind of support you’re hoping to receive in these areas.
ICYMI — ChatGPT edition
- Ithaka S+R announced a multi-year research project to assess the AI applications most likely to affect teaching, learning, and research and explore the needs of institutions and faculty members. You can sign up for their mailing list to stay on top of developments.
- This EdSurge article discusses the Department of Education’s recent report on AI and the future of teaching and learning.
- Ethan Mollick and Lilach Mollick have written a paper on seven ways AI can be used in the classroom. You can also read a summary of it on Ethan Mollick’s Substack, One Useful Thing.
- The Online Learning Consortium’s recorded May webinar on the opportunities and threats of AI in the classroom provides insights into topics such as digital literacy and data privacy.
- You can read highlights from the Chronicle’s recent virtual forum on AI and academic integrity.
Thanks for reading Teaching. If you have suggestions or ideas, please feel free to email us at beckie.supiano@chronicle.com or beth.mcmurtrie@chronicle.com.
— Beth
Learn more about our Teaching newsletter, including how to contact us, at the Teaching newsletter archive page.
