In this episode, Rob and Jared start a new series on generative AI and its practical usage in higher education. They provide a general overview of generative AI models and discuss possible ways that instructors can begin using these models in their teaching. Contact us at [email protected] with any questions. You can also visit our blog at cedarville.edu/focusblog for additional resources.

Generative AI Language models are designed to process large amounts of information and are trained to understand and generate text like a human through a vast amount of data, including books, websites, and articles.  In many ways, OpenAI's ChatGPT has become synonymous with generative AI, however, a wide variety of generative AI models have been created by different companies. Regardless of the company, it's important for the user to understand which model is being used because each of them has different capabilities. Some AI models provide the opportunity to create and train models on a certain set of information. This would enable educators to design chatbots that can assist students in learning class-specific material. An AI model can even be trained to function as a tutor which uses questions to help guide students to solutions.

Helpful next steps for instructors:

• Explore the various generative AI models (see resources below)  
• Practice prompt engineering: think about the desired outcomes, and design a foundation for the AI’s responses  
• Use a sample prompt like this one: Ask ChatGPT to use the Socratic method to train students on a specific area of your content (like writing paragraphs). Ask it to emphasize asking reflective questions, instead of giving answers. 

Resources 

• Bing Chat 
• OpenAI’s ChatGPT 
• Google Bard 
• Anthropic Claude 2 
• Llama 2