What is Generative AI?
Generative AI is a branch of artificial intelligence that can generate new content, such as text, computer code, images, or music. It uses algorithms to learn patterns from existing data and then generate novel, realistic outputs. The responses from generative AI tools, such as ChatGPT or Dall-E, can vary depending on the user input, or prompts, and the existing data the tool has trained on. While the tool may produce novel content with data it has been trained on, it will also reproduce any biases or tendencies present in the existing content. The Glossary of Artificial Intelligence Terms for Educators might help you with the terminology and various components and Generative AI Explained by AI offers an infographic version. The capabilities of generative AI are changing quickly and new tools are constantly being created. Some of these tools require payment to use; some tools may not be reliably available.
Note: more information and guidance can be found on Generative Artificial Intelligence in the Classroom.
Is it only used for writing?
No. Generative AI can be used to create a variety of media, including images, video, and music. It can also be employed to assist with language and data analysis, coding, translation, article summaries and synopsis, and literature review. As generative AI tools evolve, it is also likely that its uses and potential outputs will evolve at a similarly rapid pace. 101 Creative Ideas to Use AI in Education is a crowd sourced slide deck with examples of generative AI projects.
How are our students using it?
Students are leveraging generative AI in various ways, such as brainstorming ideas, polishing writing, automating content creation including both outlines and complete text, generating creative pieces, assisting in language translation, data analysis, and literature review. However, one of the best ways to learn how and when your students use generative AI is to ask them. How should we talk to students about Generative AI? and Conversations with Students about Generative Artificial Intelligence Tools offer some suggestions to get the conversation started.
Does Generative AI simply make things up?
Generative AI doesn’t “make things up” in the traditional sense. It generates outputs based on patterns learned from existing data. These results can improve through subsequent tries as the user refines their prompts. Some tools are connected to the internet and can incorporate results from searches in their output.
Can I ask/require students to use it?
At this time, the University of Toronto does not recommend or provide support for the use of any generative AI tool. For this reason, and taking into account a student’s potential individual privacy concerns or comfort in using these tools, U of T recommends instructors not make student use of generative AI a requirement in a course. Students should be given alternatives for any course assignment that requires the use of a generative AI tool. See Tools Beyond Quercusfor instructor considerations when considering any tool not in the University Academic Toolbox.
How can I use it for teaching and learning?
If you haven’t already experimented with a generative AI tool, now is a good time to try. (ChatGPT requires registration but has a version that is free to use.) You could start by inputting your assignments as prompts then modify based on the tool’s output. Using AI to Implement Effective Teaching Strategies in Classrooms provides five strategies and prompts to assist instructors interested in using generative AI. Let ChatGPT Be Your Teaching Assistant considers ways to generative AI tools can help lighten your teaching load.
Is there a University of Toronto policy about Generative AI use?
For information on University of Toronto policies and guidelines related to AI use and academic integrity, including Sample Syllabus Statements, visit U of T’s Academic Integrity site and the Generative Artificial Intelligence in the Classroom: FAQ. CTSI’s Generative Artificial Intelligence in the Classroom is also updated regularly as the environment surrounding AI and generative AI uses changes. Note that you have choice in whether or how you allow your students to use generative AI to complete assignments, and the University is not currently supporting the use of tools that try to detect whether writing was generated by AI. Please remember, no matter your policy on its use in a course, you should talk to your students about generative AI, its uses and implications.
How can I talk to my students about generative AI?
Instructors should be explicit about the acceptable parameters of using these tools in a course. They should be clear about their expectations, both in class discussions and on the course syllabus, and how these relate to the course learning goals. The Sample Syllabus Statement resource, produced by the Office of the Vice-Provost, Innovations in Undergraduate Education, provides valuable advice. U of T’s Academic Integrity site provides advice and guidelines for both students and instructors regarding the use of generative AI tools. Students can also consult the University of Toronto Libraries guide on citing sources for guidance on how to cite generative AI in their work.
How do I design generative AI proof assignments?
While it is impossible to design an assignment that generative AI cannot help with, you might start by considering authentic assessment design with a focus on critical thinking and analysis. Integrate library research skills into assignments to have students demonstrate their individual process of gathering sources – and ask your liaison librarian for support. Make assignments specific to material covered in classes and tutorials, including class discussions and student presentations. Scaffold assignments to build on work already produced by the student. Focus more on the process than the product and include a reflective component. Consider a variety of assessments, such as presentations, moderated group discussions, or a portfolio. Provide a rubric or grading criteria material to ensure students are well aware of expectations. Be sure to check in with your students regularly to gauge progress and if they have concerns or questions on the process. For more information on this topic, visit CTSI’s Generative Artificial Intelligence in the Classroom.
What are three things I should do now ?
- Talk about it with your students. Discuss your expectations and guidelines around using generative AI tools. Learn how and why your students use them.
- Rethink your assessments with use of AI in mind.
- Try it to get a better understanding of its capabilities. Try it yourself and talk to your colleagues, consult with a teaching centre faculty liaison or librarian. Remember you are not in this alone—we can all share and learn with each other.
Note: The first draft of this document was created using ChatGPT. We provided the ten prompts then modified and updated the content (including links) for the U of T context.