网曝吃瓜

On April 3^rd, Sam Jay held a 3-hour workshop titled AI-Aware Assessment. I presented in the first hour and tried to set the stage for the really important and really difficult work we have ahead of us. We had roughly 60 participants in the room, with well over 100 watching remotely.

As part of that session I polled the participants using . Between 70 and 80 people responded to the four questions I asked. These aren鈥檛 questions I have asked before and I was surprised by the responses in each case. I will share each question, the results and my thoughts.

Question 1:

鈥淕enAI is inherently problematic and use of it should be avoided at nearly any cost.鈥� Agree or Disagree?

77 responses: 35%-Strongly Disagree, 36%-Moderately Disagree, 9%-Neutral, 12%-Moderately Agree, 8%-Strongly Agree.

Looking across the social circles in my own I鈥檓 aware that different individuals and groups have very different perspectives on GenAI. I think I expected the variety of faculty represented by these responses would be more widely distributed.

I had two goals in asking this question. First, I genuinely didn鈥檛 know how this (self-selected) group of MSU Dever faculty and administrators would answer. More importantly, I don鈥檛 think we all know what each other thinks. This kind of result can help us orient our attention and help us take action. Second, I really struggle with my own answer to this question, while at the same time I think it is essentially moot: Whether I believe is inherently problematic or not, generative AI is already impacting higher education in ways that demand action and attention. I believe that in our work at 网曝吃瓜 we have to grapple with the widespread use of these tools, regardless of their ethical qualities.

I hope I don鈥檛 look back on that statement with deep regret later. For now, I think we have to move ahead.

Question 2:
鈥淚 believe there will be software/tools that will allow us to keep using the same asynchronous assessments we were using 5 years ago.鈥� Agree or Disagree?

70 responses: 57%-Strongly Disagree, 24%-Moderately Disagree, 10%-Neutral, 6%-Moderately Agree, 3%-Strongly Agree

The arms race around 鈥渃heating as a service鈥� (CAAS, if you will) is not new. Professors have used search engines to see if student writing was plagiarized from Wikipedia. Websites like Chegg would give a huge metaphorical wink to learners, clearly promising fast, easy answers for a low monthly price. And now, that arms race has escalated again.

Since ChatGPT landed in November 2022 I鈥檝e seen worried, passionate, exhausted faculty ask, 鈥淐an鈥檛 we just buy a detector to catch students?鈥� To me, the answer has been a clear 鈥淣o鈥� for a while now, and I am gratified to see such a strong majority of responses pointing in that direction.

The truth is that our assessment methods (especially asynchronous assessments) have had huge problems for years or decades. GenAI has taken those cracks and blown the whole thing wide open. Assessments done asynchronously (regardless of course format) must change.

Question 3:
Looking ahead 5+ years, rank these possible futures:

1. GenAI is less present in human life than it is in 2026
2. GenAI is peripherally involved in many aspects of human life
3. Gen AI augments many aspects of human life in important ways
4. GenAI is fully integrated into nearly all aspects of human life
5. GenAI (or AGI) dominates most or all aspects of human life

The 74 responses yielded the following results: Option C (GenAI augments many aspects鈥�) dominated the rankings, while option A (GenAI is less present鈥�) was clearly last. The other three were roughly tied in the middle of the pack.

Much like the question about the fundamental morals of generative AI tools, I didn鈥檛 know how this group of people would answer this question. I find that I am very skeptical of the claims made about Artificial General Intelligence (AGI), but I worry that my own opinion is self-serving and heavily rationalized. Regardless, it was reassuring to see these results pointing to an expected future that heavily involves GenAI. 48 of the 74 responses (65%) ranked 鈥淕enAI is less present鈥︹�� dead last.

My takeaway here is that we have surprisingly broad agreement about the importance of GenAI. No doubt it will show up in very different ways across our subjects and courses, but in general we think our students will be using these tools a lot.

Question 4:
Who should take primary responsibility for adjusting to how GenAI is impacting higher education?

1. Faculty
2. Students
3. Staff (instructional designers, advisors, tutors, etc.)
4. Academic Administrators (Chairs, Deans, AVPs, etc.)

75 responses: 1%-Students, 3%-Staff, 61%-Faculty, 35%-Academic Administrators.

Even now, weeks later, I can鈥檛 quite decide if this is what I expected or if it is a surprise. I think I would have predicted more votes for 鈥淪taff鈥� but otherwise this makes sense. My was to use this question and the classic to lead into the lack of action. I said that perhaps鈥�

• A faculty member looks at their dean and thinks 鈥淚 don鈥檛 see them working on this, so it must not be that important.鈥�
• A department chair looks at their senior faculty and thinks 鈥淭hey aren鈥檛 unified on the need to redesign curriculum, so things must be okay.鈥�
• A Staff Senator looks at the Faculty Senate minutes for this year and thinks鈥� The faculty senate hasn鈥檛 voted any kind of GenAI policy for either student or faculty use. No need for us to work on our policy.鈥�
• An AVP listens to department chairs and thinks 鈥淐hairs are more focused on budgets and enrollments, so those are clearly more important.鈥�

I have not heard any of these statements from real people at 网曝吃瓜. But when I consider how the reality of this problem is so much bigger than the scope of our response, this is one story I make up to explain why.

The only way past is through

During my session I said that ChatGPT 鈥済raduates鈥� this spring. At spring commencement in May, there could be students who used generative AI in every single course they took during their 4 years at 网曝吃瓜 (fall 2022 through spring 2026). Indeed, I think there must be students where most faculty, staff and administrators would agree, 鈥淣o, wait. You didn鈥檛 actually earn these passing grades. You don鈥檛 have the education that an 网曝吃瓜 diploma is supposed to certify.鈥�

Meeting this moment is going to be hard. No one part of the university can do this. No expert will appear with 鈥渂est practices鈥� backed by 10 years of evidence. The only way through this is with active listening, steady communication, urgent transparent action, and participation across all levels.

Notes and Disclosures

The featured image is by Pawel Czerwinski and is free to use (via Unsplash).

Generative AI disclosure: After writing this piece I used generative AI to write a first draft of the short 鈥渢easer blurb鈥� that went out by email. Want to know more? Send me an email and we can chat!