Module 1: Introduction to Generative AI
This module provides educators with a foundational understanding of GenAI, empowering them to acknowledge GenAI’s presence in higher education and make informed decisions about mindfully integrating these tools into their teaching practices.
Objectives
- Define generative AI: explore what GenAI is and its role in modern technology.
- Name some common GenAI tools in higher education for teaching and learning.
- Discuss the potential applications and limitations of GenAI.
I. What is GenAI?
First key questions:
- What images or ideas come to mind when you hear the term artificial intelligence?
- What tools pop up first when you think about generative AI?
The term artificial intelligence often conjures images of robots and high-tech scenarios, but its academic definition is rooted deeply in history. Coined by John McCarthy in 1955, artificial intelligence was envisioned as the science of making machines capable of performing tasks that would require intelligence if done by humans (Calo, 2017). Today, this encompasses a wide range of technologies, but it’s generative AI—or GenAI—that’s sparking significant interest in educational spheres.
GenAI is a type of artificial intelligence (AI) that is able to create new content, such as text, images, music, or entire datasets, based on patterns and information it has learned from existing data. Unlike traditional AI that simply analyzes data, GenAI actively produces new material, simulating a level of creativity once thought unique to humans.
More readings:
- Calo, R. (2017). Artificial Intelligence Policy: Primer and Roadmap. U.C. Davis Law Review, 51(2), 399-436.
- McCarthy, J., et al. (1955, August 31). A proposal for the Dartmouth summer research project on artificial intelligence.
Activity 1: Watch and reflect
Watch: Introduction to Generative AI from Google Cloud Tech. This 22-minute video provides a primer on how GenAI operates, including its applications and basic model types.
Reflect: How does GenAI work? How does GenAI differentiate from other AI technologies? In what ways could you envision utilizing GenAI within your own teaching or administrative context?
Activity 2: Read and reflect
2.1. Dive into AI Explained, a beginner’s guide, which offers a succinct glossary of AI terms and concepts
Reflect: How might you describe supervised learning, unsupervised learning, and reinforcement learning in the context of machine learning? Do you see any interesting parallels between machine learning and human learning?
2.2. Explore A curious person’s guide to artificial intelligence . This article provides another overview giving you just enough of a foundation to talk about AI and get thinking about how you would like to explore it.
Reflect: How do chatbots like ChatGPT and Bard interact with users, and what are their limitations when it comes to interpreting meanings?
II. GenAI capabilities and limitations
AI is not new. In fact, it has long been a part of our daily digital interactions from curating social media feeds to recommending products or routes and booking flights. But why has there suddenly been a new wave of interest in the field since the release of ChatGPT 3.5 in November 2022? Millions of people tried this tool within a month and, since then, several AI platforms have emerged almost daily.
As Bowen and Watson (2024) suggest, while previous AI primarily curated the world, generative pre-trained transformer or generated pre-training (GPT) AI actually has the potential to allow us to create the world. Though this is only the beginning of understanding what AI can be capable of, it’s critical to understand its current capabilities and limitations as of now.
Activity 1: Engage with a chatbot and reflect on its responses.
Objective:
Explore the understanding and communicative capabilities of GenAI through interaction with some GenAI tools and critically analyze their responses.
Tools:
Choose from AI chatbots such as Bing, Gemini, Claude, or ChatGPT.
1.1 Task:
Interaction:
- Prompt: Ask the selected chatbot, “What are some potentials and limitations of GenAI in higher education?”
- Process: Engage in a detailed conversation to probe deeper into any responses given. You may ask follow-up questions to clarify points or explore related topics.
1.2 Reflection:
After your interaction, critically reflect on the information provided by the chatbot.
- How comprehensive and accurate were the responses?
- Were there any notable insights or unique perspectives offered by the chatbot?
- Identify any areas where the chatbot’s responses seemed limited or biased.
- How do the chatbot’s responses align with or challenge your current understanding of GenAI’s potentials and limitations in higher education?
Activity 2
Read this book chapter on the capabilities and limitations of GenAI.
Activity 3: Read the following summary of the capabilities and limitations of GenAI. Reflect on the first three steps that you would need to do if you plan to integrate GenAI into your coursework.
While the potentials of GenAI are numerous — from tailoring educational content to unique learner profiles to breaking language barriers — it is not without its challenges. Concerns about data privacy, the perpetuation of biases, and the potential for academic dishonesty are paramount. Furthermore, the environmental impact of training large AI models and the ethical considerations around AI-generated content need thorough scrutiny.
GenAI capabilities:
- Creating content: they can write various forms of text like stories, poems, and code.
- Generating educational materials: these tools can produce examples, outlines, and even long texts like essays or reports.
- Summarizing and feedback: they can summarize text and give feedback on the writing’s structure, style, and grammar.
- Teaching aid: they explain ideas at different levels, making them great for teaching.
- Language tools: they can translate text between languages.
- Memory: some tools can remember previous instructions during a conversation or from past interactions.
- Media creation: certain AI tools can also create images and videos.
GenAI limitations:
- Bias: AI may reflect biases from the data it was trained on, potentially leading to unfair or discriminatory results.
- Errors (“hallucination”): AI sometimes makes things up or gets facts wrong, which is referred to as hallucination.
- Environmental and human costs: running these AI models requires a lot of energy, which can have environmental impacts. Some AI development practices have also raised concerns about the well-being of the workers involved.
- Misinformation: AI can generate realistic, but fake, content, which might spread misinformation, especially via social media.
- Copyright issues: AI tools use vast amounts of online data, including content that may not be freely available or intended for such use, which raises legal concerns.
Privacy and safety: without strong regulations, the use of AI can pose privacy and safety risks as these tools often collect and use large amounts of data.
Activity 4: Group discussion
Explore the potential challenges and opportunities presented by GenAI in academic settings through this “what if” scenario-based table. This exercise, developed in collaboration with ChatGPT 4, is designed to initiate discussions and encourage critical thinking among peers and students regarding the impact of GenAI in teaching and learning.
| Scenario | What if? | Opportunity | Challenge |
|---|---|---|---|
| Personalized learning | GenAI could automatically tailor all educational content to each student’s learning needs and pace in real time. | ||
| Automated research | GenAI tools could autonomously conduct literature reviews and data analysis. | ||
| Content creation | GenAI could create educational materials, such as textbooks, lectures, and exams in all fields. | ||
| Language translation | GenAI could instantly translate educational materials and lectures into multiple languages. | ||
| Academic integrity | GenAI tools could generate entire essays or research papers on demand. | ||
| Real-time exam monitoring | GenAI could monitor students during exams to detect cheating using facial recognition and predictive behaviors. | ||
| Bias in educational AI | GenAI systems inadvertently reinforce biases present in their training data within educational settings. | ||
| AI as tutors | GenAI could function as virtual tutors or even instructors for all subjects. |
Summary
In this module, we explored the definition of GenAI and gained some insights into its presence in teaching and learning contexts. We also briefly examined how these tools can be applied in an educational context while considering their limitations.
Final reflective question
Reflect on your recent teaching session. Consider whether and how GenAI tools could have enhanced the learning experience and outcomes. What concerns might you have about integrating such technologies into your teaching practice? Does the integration of GenAI align with your educational values and goals?
