What is generative AI?
The most prominent Generative AI tool people are discussing in higher education is ChatGPT, a product from OpenAI. Generative AI tools can create new data – like text and images (see below) – based on patterns it has learned from the large sets of existing data on which it is trained. ChatGPT is a chatbot tool intended to generate text in response to requests and input from users. Tools have also been developed to generate other types of data, such as Adobe Firefly & OpenAI’s DALL-E 2, which can produce images based on short strings of text provided by users, and Soundful.com, which is AI generated audio. Other technologies, including generative AI video, seem to appear every day.
While Google, Microsoft and other companies have featured AI in many tools and platforms, they, along with companies like Adobe, are beginning to integrate generative AI into search and productivity tools. For example, Microsoft now has integrated OpenAI's GPT system into its Bing search (use the "Chat" mode in Bing in the Microsoft Edge browser) and is now working on integrating generative AI tools into its productivity software, such as Word and Powerpoint. Google recently made its Bard AI chatbot openly available and is planning to soon launch AI-powered tools in products like Gmail and Google Docs.
Ok, but what does all this mean for higher education?
This moment in time is still something of the "Wild West" when it comes to Generative AI and higher education. The truth is there are probably more questions than answers at this point. Some common questions or concerns from faculty include:
- Can AI-generated output can be passed off as human generated? If so, how will I know?
- Do we have any policies on Generative AI? If so, how do we enforce them?
- Given its potential, how might we address Generative AI in our learning outcomes, activities, and assessments?
- If students or faculty are using these tools, can they be trusted? Can their content be trusted? What is the likelihood for it to produce biased output?
- How might these tools aid in teaching and learning? Are there any best practices or guides available?
- How are these tools impacting our own disciplines, and what we can do to help prepare students for a world where skills in using Generative AI will be of tremendous value.
How does it work?
Before we get into integrations and implementations or even policies and practices, let's start with a bit of a hands-on activity just to make sure we are all familiar with the tool:
Activity 1 - Playing with ChatGPT
Step 1 (1 min): Choose a common writing genre or document form from your field and identify a key topics or issues of interest in your discipline.
- Example: Genre/form - Abstract (conference presentation abstract). Key topic: Generative AI's impact on writing courses.
Step 2 (4 min): Go to ChatGPT and have it create a draft of your genre on your topic.
- Prompt given to ChatGPT: Write a 500 word conference proposal for a presentation exploring how Generative AI is changing writing courses in higher education
Step 3 (3 min): Partner up or in small groups, share what you learned/experienced from this exercise.
Step 4 (3 min): Share Out (i.e., volunteers to share conversational insights with the group)
Uses in Teaching and Learning
It's all in the prompt. There are lots of strategies for using ChatGPT (and other text generative AI tools in higher education). Ethan Mollick, a UPenn Wharton School professor has written quite a bit in recent months about generative AI and some of the examples below are adapted from his work. In particular, you may want to check out his posts "Using AI to make teaching easier & more impactful" and "How to use AI to do practical stuff: A new guide." More recently, he's written about how assignments can be designed to challenge students to learn the technology.
Create formative assessments
One of Mollick's prompts includes using ChatGPT to create low-stakes formative assessments to add to courses. His example instructions for ChatGPT is below:
Create hypotheticals, examples, and question variations
In many of classes, instructors use hypothetical examples to ask students to apply course concepts. Those examples can be time-consuming and difficult to develop. This is also the case when instructors want to create multiple versions of a question.
Example Prompt for ChatGPT: When learners encounter new ideas and concepts, especially complex ideas and concepts, having different examples can help them learn better. I want you to act like an example generator and create 3 different examples of X in action for Y-level students. e.g. X = ethos; Y = Freshman College Students. (Mollick has a similar prompt).
You can also provide existing test questions and ask it to create variations. And keep tweaking until it's to your liking.
Write, revise, and edit
Generative AI tools can be used to write or help you write. 3 Quick Ways it functions in and around writing are:
- Getting Started: Some see it as a draft creator and an easy place to get the writing started. It can help with writer's block by eliminating the paralysis of the blank page.
- Peer Reviewer & Copy Editor: You can ask these tools to revise writing to make it more concise. Or to adapt tone for another kind of audience.
- Rhetorical Invention: Tools like ChatGPT can also be used to come up with ideas, a lot of them, quickly. Most are just okay as far as ideas go, but there may occasionally be a diamond in there. Sometimes seeing a large set of ideas can also help with sparking one's own creativity.
It'll take a little work, but these can be incredibly useful in writing or other creative processes.
What other ideas do you have for ways in which these tools can be used effectively in teaching and learning?
Generative AI & Syllabi: An Ethics of Practice
Excerpted from IUB COAS Policy Guidelines
To Instructors: If your syllabus does not currently include an academic integrity statement, we encourage you to incorporate the text below (feel free to modify the statement as needed). In addition, if there are specific policies that apply to your course (e.g., regarding group work, collaborative assignments, citation requirements for textbooks and/or class notes, etc.), these should be included in your statement.
Statement on Generative AI: According to the Code of Student Rights, Responsibilities, and Conduct, plagiarism is “presenting someone else’s work, including the work of other students, as the submitting student’s own. A student must not present ideas or materials taken from another source for either written or oral use without fully acknowledging the source, unless the information is common knowledge.” Cheating is “using, providing, or attempting to use or provide unauthorized assistance, materials, information, or study aids in any form.” (II: Responsibilities B,4,a and c) Your use of anyone else’s (or any AI system’s) words, graphic images, calculations, or ideas should be properly cited. AI generators/programs such as ChatGPT, Grammarly, QuillBot, Spinbot, Dall-E, etc. should not be used for any work for this class without explicit permission of the instructor and appropriate attribution. AI text generators should not be used for:
- Creating or revising drafts
- Editing your work
- Reviewing a peer's work
The use of generative AI platforms will be considered plagiarism and/or cheating and will be reported to the Dean of Students (Office of Student Conduct) and handled according to University policies. Sanctions for academic misconduct in this course may include a failing grade on the assignment, a reduction in your final course grade, or a failing grade in the course, among other possibilities. Contact your instructor if you have questions.
Generative AI in professions, including medicine
In the past few months, many have written about the impact of generative AI on various industries, with McKinsey & Company estimating that it will have the ability to automate nearly 70 percent of what employees currently spend time doing and adding hundreds of billions of dollars to the economy. Healthcare, as one example, is no exception to the disruptive force, with providers now using the technology to do things like summarize research or write case summaries. One health system has recently penned a deal with Microsoft and EMR company EPIC to integrate the tech into EPIC's MyChart system to help providers draft messages to patients. In fact, some physicians have said the technology has helped them improve communication.
There are many potential uses of generative AI in healthcare and medicine, and we're seeing increased discussion of necessary ethical considerations as well.
What ways do you envision generative AI impacting education in your disciplines? What are the opportunities, challenges?
What ways do you envision generative AI impacting the professions your students are likely to enter? What are the opportunities, challenges?
Resources
There is much being written about these tools, including its role in teaching and learning. Adam curated a few items below.
- IU Knowledge Base: Acceptable uses of generative AI services at IU. This document provides an overview of precautions you should take when using generative AI tools at IU (especially around the type of data shared with the services). It also includes an overview of potential acceptable uses for generative AI tools at IU. (Note, "acceptable uses" refer to what is acceptable from a data sharing/security perspective; what is ethically acceptable is a much more complicated discussion bounded by evolving disciplinary/organizational norms.)
- IU Knowledge Base: Precautions about using ChatGPT at IU. This includes important information about the use of generative AI at IU, including NOT sharing institutional data with the tool and being knowledgeable and transparent about the data collected by the tool.
- IU Connected Professor: Generative AI in EDU: How tools like ChatGPT are affecting teaching and learning.
- IUB's Center for Innovative Teaching and Learning's page on productively addresses AI-generated text is teaching.
- ChatGPT and meaningful writing: How to work with AI through assignment design — A presentation from CITL's John Paul Kanwit and Sean Sidky
The following are articles, resources, etc. about this emerging technology, curated by IU's own Emily Oakes and shared with other universities that are members of Unizin (Unizin is a multi-university consortium IU of which IU is a founding member.)
- Creating more effective assignments
- Sample ChatGPT prompts
- Sample syllabus/classroom policy statements
- Ways to use generative AI tools in teaching and learning
Additional articles
- ChatGPT is fun, but not an author
- Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models
- ChatGPT and the Rise of Large Language Models: The New AI-Driven Infodemic Threat in Public Health
- Chatting about ChatGPT: how may AI and GPT impact academia and libraries?
- ChatGPT: Bullshit spewer or the end of traditional assessments in higher education?
- What ChatGPT Can’t Teach My Writing Students
Credits:
Created with images by HappyBall3692 - "Businessman analyzing growing virtual hologram data. Concepts of technology, finance, business and investment. Stock market and digital funds Business assets that analyze charts, trading, " • HappyBall3692 - "Businessman analyzing growing virtual hologram data. Concepts of technology, finance, business and investment. Stock market and digital funds Business assets that analyze charts, trading, " • HappyBall3692 - "Businessman analyzing growing virtual hologram data. Concepts of technology, finance, business and investment. Stock market and digital funds Business assets that analyze charts, trading, " • HappyBall3692 - "Businessman analyzing growing virtual hologram data. Concepts of technology, finance, business and investment. Stock market and digital funds Business assets that analyze charts, trading, " • HappyBall3692 - "Businessman analyzing growing virtual hologram data. Concepts of technology, finance, business and investment. Stock market and digital funds Business assets that analyze charts, trading, "