How We Use AI In Schools
One of my professors reached out to me about his AI Policy that he’s going to include for next semester.
He wrote:
Al tools like ChatGPT can be helpful for generating ideas, enhancing understanding, or getting feedback on your writing. However, they should not be used to complete assignments for you. All submitted work must be your original creation. You are expected to fully understand and critically engage with any Al-generated content, and failure to properly acknowledge Al assistance could result in academic integrity violations. Always follow specific guidelines from the instructor regarding Al use. (This paragraph was written by Al.) From Tom: As evidenced above, AI can do a good job of forming text when given the right prompts/questions. What AI cannot do, however, is learn for you. Yes, you can learn from what AI has generated, but that is different than you constructing knowledge. The old adage “Good things don’t happen without effort” applies here. What you gain derives from what you invested, including your time, thinking, and, yes, some angst.
I really appreciate that he is thinking about this and is bringing it up for feedback.
I responded as follows:
Your paragraph is much more powerful:
From Tom: As evidenced above, AI can do a good job of forming text when given the right prompts/questions. What AI cannot do, however, is learn for you. Yes, you can learn from what AI has generated, but that is different than you constructing knowledge. The old adage “Good things don’t happen without effort” applies here. What you gain derives from what you invested, including your time, thinking, and, yes, some angst.
That’s a much better thing to say than what AI wrote. You should know that our current professor STRONGLY encourages the use of Grammarly, which is just an earlier version of technology-assisted writing.
That’s what AI is offering: technology-assisted writing.
This AI policy paragraph has two major problems, which I’ll enumerate below:
Al tools like ChatGPT can be helpful for generating ideas, enhancing understanding, or getting feedback on your writing. However, they should not be used to complete assignments for you. All submitted work must be your original creation. You are expected to fully understand and critically engage with any Al-generated content, and failure to properly acknowledge Al assistance could result in academic integrity violations. Always follow specific guidelines from the instructor regarding Al use. (This paragraph was written by Al.)
- The hypocrisy of professors and teachers to use AI for their work but deny students the ability to use it is astounding.
- “Failure to properly acknowledge[^1] AI assistance” is a blatant, disgusting threat, loosely termed to fully benefit the professor and open the door for spiteful behavior (i.e., if I like you, it’s fine, but if I don’t, I’ll accuse you of academic integrity violations). And it is downright insulting when our classes have largely been disorganized repeats or made up on the fly. Professors have literally dumped previous courses directly into Canvas and while it looks brand new to us, it’s just a repeat of previous courses (or they have not known what they are doing and haven’t provided a clear outline.)
Here’s my suggestion for an AI policy: AI should enhance learning, not circumvent it. You said it very well in your closing paragraph in your email.
The very real problem is our assignments are hoop jumping assignments. They don’t mean anything. They don’t matter. For example, our qualitative research class right now is having us do qualitative coding by watching a movie and coding that movie. It’s ridiculous, and a waste of time because it is not real. A much better way to do a qualitative research course would be to have us actually do a qualitative study and submit it for publication. If I were teaching this qualitative research class, that would be my goal, have everyone complete a qualitative research study and submit it for publication.
Here’s a crazy idea for your class: Let’s draft a book together next semester. Commit to publishing a book from our writing. Then, our assignments become real and have actual value. We can write it in Office365 online or Google Docs and we will have version history to see when people are making additions and changes. We can commit to disclosing how we used AI, if that makes everyone feel better.
Or, we could each write an article for your multiple intelligences newsletter. Or several blog posts for the Center. There are lots of real-world options out there.
Honestly, submitting a paper or completing a chapter for our dissertation should really be the focus of each of these classes we are taking.
In the absence of valuable assignments, we are going to do what any sane person would do who has a lot to do and little time to complete it: we are going to turn to the tools that make it as easy and fast as possible for us to get work done.
When the assignments are not real, and the professor is the singular audience, it makes total sense to get it done as quickly and simply as possible, thus folks turn to AI, because it is good enough. And most of the time, it is good enough.
When we care, we don’t use AI because we are probably better writers without it. When we don’t care, and we know we are getting slop, we naturally turn in slop.
This program has not done a good job of appearing professional and high integrity itself, as mentioned by some examples above. Was this complicated by at least a dozen reasons, most of which are outside of your control? Yes. Absolutely. Surely, the unexpected death of Dr. Bell and other life circumstances and challenges played their part.
This is still our only (hopefully) opportunity for a doctoral degree. Surely none of us want to delay or transfer to another program. We have started, we are committed, and we want to finish.
In closing, and to be clear, I am speaking for myself, not for the rest of the cohort here. I don’t know how they would feel about the content of this email.
Tom, I respect you and appreciate you taking the time to hear me out about this. The fact that you even reached out to begin with means a lot to me. Thank you for being an example and a leader worth following. [^1]: I used AI to finish spelling acknowledge by pressing Tab on my computer as I was typing. Do I need to acknowledge that? And what is the proper way to acknowledge it? These are questions that I’m certain you don’t want to waste your time arguing about.