These are open questions! As you ask them, keep an open mind, and be ready to go “off script.” Your students’ answers may surprise you.
This, the broadest of all GenAI ethical questions, might have many dimensions and is difficult to tackle without devolving into broad generalizations. It might be useful to start this conversation locally, examining possible impacts right here in the university environment.
GenAI and Learning
To begin with, what impact might GenAI have on learning here in our classes?
- Most students agree that using GenAI to complete entire assignments is like bringing a forklift to the gym (see Chiang, 2024)—you won’t build the muscle you need that way. But where should we draw the line between AI use that supplants learning, and AI use that might support learning?
- Even if most students are not using GenAI for “cheating” most of the time, how might the general public come to believe that “all students are cheating with GenAI,” and what kind of consequences might that broader belief have?
- What kind of differences do you see between “using GenAI in college” and “using GenAI at work”?
- What kind of expectations do you think your professors have for GenAI use, and how do they differ (if they do) from what your future supervisor or boss will expect?
GenAI and Teaching
Sometimes, expectations are also lopsided. Students are often acutely aware of this.
- How many of your professors say they use GenAI in preparing course materials for you to use?
- How many of your professors do you think are using GenAI in preparing to teach, but don’t disclose it?
- How comfortable are you with these ideas about professors’ use of GenAI?
- Sample Resource: Students have sued universities over undisclosed use of GenAI in teaching.
- We also have evidence that academics are using GenAI in their research and especially writing, without fully disclosing it. How do you feel about that?
- Sample Resource: An Inside Higher Ed article suggests that the use of GenAI in research writing is rampant and widely underreported.
GenAI and Authenticity
Even with clear expectations, there may be limits to how well we can ensure everyone meets them.
- How much do you know about others’ use of complex digital technologies?
- How would you compare how your peers or colleagues use GenAI with how they use social media or cellphones?
- Can you tell a real photo from an AI generated photo? If not, how much is the admonishment “pics or it didn’t happen” still useful?
- Sample Resource: Test yourself! This online test lets you compare “real” and AI-generated photos side by side. How well did you do?
- Sample Resource: A New York Times article interviewed a man who professionally investigates cases of deepfakery.
- If it turned out that one of your favorite movies, books, or songs was mostly created using GenAI, how would that change your opinion of it?
- GenAI LLMs were trained mostly on the English-language Internet. That means that they do not include quite a large amount of information—they don’t represent the world equitably. How might that influence your learning here at UTK? Or, if “things are different” here in college, how might these mismatches between the LLM and the world play out in the wider culture here at Knoxville, or in your hometown?
GenAI and Reality
The technology to create “deepfakes” is becoming increasingly publicly available, and it’s an open question how society at large should respond.
- If you saw a shocking photo purportedly taken right here on campus, how would you verify if it’s real? What if it’s very recent, and couldn’t be assumed to have reached the national news?
- How do you judge how real day-to-day news is? What are your go-to media channels, and how much do you trust them?
- What kinds of rules or regulations might mitigate concerns about “deepfakes”? How well do you think they would work?
Privacy and Security
Privacy and security as such have been at greater risk ever since the introduction of the Internet itself, and complex digital technologies (not just AI) have created new complications and risks.
- How often do you read the “Terms and Conditions” you have to accept to use a website or piece of software?
- It can be a fun exercise to explore some of these together in class.
- How familiar are you with the “Privacy Policy” of the social media or other software you use regularly?
- Again, it can be a fun exercise to explore some of these together in class.
- What are “cookies,” and why do so many websites ask you what you want done with them?
- How do you tend to respond?
- Have you noticed which websites allow you more (or less) control?
- If corporations are harvesting your data for financial gain, how much of a cut (if any) do you feel they owe you, and why (or why not)?
- Here at UT, we have UTVerse and the AI Hub, which do not allow your session data to be shared outside the UT digital ecosystem. How much do you know about these systems, or how they work? Do they give you “enough” privacy? Why or why not?
For Further Thought and Discussion
At the broadest level, it’s worth exploring all the ways in which complex digital technologies can influence society and culture, and to ask these questions each time we find a new use case. This was the purpose that Jon Ippolito had in mind when he created the IMPACT RISK framework.
Sample Resource: The IMPACT RISK framework, created by Jon Ippolito, explores a number of possible negative impacts that GenAI could have on society and culture.