By Doug Ward
The phrase “humans in the loop" has become a cliché for the importance of overseeing the processes and output of generative artificial intelligence.
The rapid changes that generative AI have brought about, though, often make us feel like we are caught in an endless digital loop. Since the release of ChatGPT 3.5 three years ago, a bombardment of announcements and changes have made it hard to cut through the noise and gain clarity about the direction of this new AI-fueled world. ChatGPT and competing AI models have improved with head-spinning speed, new tools have been released almost daily, and those tools often blur the lines between the real and the artificial.
The music video above is my tongue-in-cheek commentary on the digital surrealism of the past three years.
I think back to what an instructor said during a workshop nearly three years ago: “I just wish someone would tell us what we are supposed to do.”
That wasn’t a plea for a mandate. It was an expression of exasperation of how education had been turned upside down, with no clear way to adapt.
I don’t see the changes slowing anytime soon. The only way to move through the maelstrom is to ground ourselves in core principles, embrace the pace of rapid change, and adapt our methods of teaching and learning. The good news is that the same AI tools that have challenged existing approaches to education can also empower us to rethink and remake learning for a generative world.
At CTE, that has been our message from the start. We may feel like humans in a loop, but our students need us to stop spinning and push through the maelstrom.
Use the break to experiment
Here’s a challenge as the semester winds down: Use winter break to introduce yourself (or re-introduce yourself) to generative artificial intelligence. Experiment with at least one generative artificial intelligence tool and find a way to integrate it into an assignment in the spring.
If you already feel comfortable with generative AI, experiment with a new tool and consider how you might use it to rethink instruction, create new approaches to online and in-person learning, enhance student skills, and deepen student understanding.
If you aren’t sure where to start, here are some ideas:
Copilot agents
Copilot agents allow you to give custom instructions and information to Copilot and create a personalized bot for your work or for your class. I’ve written previously about how to create Copilot agents. I’ve also created three agents to help with using Copilot. You’ll need to log in to Copilot with your KU credentials to use them.
- Copilot Agent Idea Guide. This tool will help you learn about Copilot agents, provide examples on how you might use them, and guide you through creation of your own agent.
- Agent Idea Generator. This is much like the Idea Guide, but it is intended to help students consider ways to use Copilot for learning. I created it for my class this semester, and students said they found it helpful.
- Rubric Assistant. This will help you create a rubric or help you improve an existing rubric.
NotebookLM
NotebookLM allows you to compile articles, links, videos, notes, and audio files, and use Google Gemini to explore those materials. Chat results in NotebookLM include links to passages in your material, allowing you to check the accuracy. NotebookLM also allows you to create concept maps, infographics, video and audio overviews, slide decks, flashcards, and quizzes from your materials.
Here’s a notebook I’ve called Uses of AI in Learning. It includes examples of assignments other instructors have used and ideas on other learning activities to try with generative AI tools.
The free version of NotebookLM is fairly generous. You can create up to 100 notes with up to 50 sources in each. Those sources can be up to 500,000 words or 200 megabytes. You can have 50 queries a day with your notebooks and create three audio overviews.
Doug Ward is associate director of the Center for Teaching Excellence and an associate professor of journalism and mass communications.