What are the most effective ways to use artificial intelligence in cooperative learning?

Using AI in group learning is my thing, and here's how it helps a lot: Smart Groups: AI picks the best mix of people for groups based on what they're good at. Personal Learning: AI makes lessons just for you, based on what you like and how you learn. Quick Feedback: AI tells you how you're doing right away so you can get better. Predicting Problems: AI sees if there might be issues in the group and helps fix them early. Chatbots Help: Talking robots (chatbots) help answer questions and keep group chats going. Learning Stats: AI looks at how groups work and helps make learning even better. Adding AI to group learning makes it easier and more fun for everyone!

Equipped with the right input, AI can provide adaptive and personalized support to each group member based on their learning needs and progress. This will lead the way to facilitate peer feedback and assessment by providing rubrics, criteria, and examples.

At Kennesaw State University, our research team is developing and piloting "rhetorical shot" prompt engineering to help students engage with Ai-assistants in collaborative content development and peer review. We have also surveyed more than 1,200 first-year students and gauged their sentiments towards AI-infused assignments in their courses. Our goal is to develop effective "rhetorical shot" methodology to increase student writing success in their academic and workplace writing contexts.

One thing we discovered about AI acting as a tutor is that it provides personalised and adaptive instruction to each student based on their learning needs, preferences, and goals. It can also assess the students’ learning outcomes and provide feedback and guidance. Implementing the right strategy including tools like AI, can help students to master the content and skills that are necessary for effective cooperation.

Generative AI assistants, especially Claude AI, have a unique ability to interact with student writing and help students revise their writing as a human tutor would. For example, students draft their content, and they are able to recursively revise it for audiences, context, style, and tone by prompting their AI assistant.

Interestingly, AI can act as a peer that participates in the group work as a co-learner, co-creator, or co-evaluator. It can also model and demonstrate cooperative behaviours and strategies, such as sharing, listening, questioning, explaining, and reflecting. Sure, we need more data to avoid bias, but overall it might help things kick off when we feel stuck in our creative process.

Out of the 1,200 college students we surveyed, more than half of them consider AI assistants to be positive additions to peer learning. While only 30% of these students responded that they use AI in their academic writing, they also have a generally positive sentiment in their qualitative responses. This leads us to think that college students consider generative AI as a value-add for their writing.

As a professor, I have used generative AI to help me design writing assignments. For example, if I want to design a Rogerian argument assignment that is transparent and understandable to a student audience, I can revise my prompt shots to include the five rhetorical elements: (1) purpose; (2) audience; (3) genre; (4) style, and (5) social context. I can use my model of rhetorical engineering to revise and revisit each of these five elements to get to the most relevant assignment parameters with the most effective learning outcomes.

A key advantage of generative AI is that it learns when humans collaborate with it and input information. When we interact with generative AI, it sometimes gives the wrong information. As we interact with AI models, our responses to them, in fact correcting them, particularly on non-generalized knowledge, can improve their accuracy and encourage human participation in LLM training.

When integrating AI into cooperative learning, think of it as a dynamic ecosystem member. AI’s effectiveness is maximized when it adapts to and from the learning environment—structuring groups, personalizing tasks, and evolving through interactions. This creates a reciprocal learning relationship, where AI and students learn from each other, fostering a continuously improving educational experience.

Prudence needs to be exercised while using AI. Data it is using to teach itself on, results itbis generating, contributions itnis making as a lesrner, facilitator or a peer- all needs to be closely monitored.