Frequently Asked Questions on Using ChatGPT in the Classroom
After two years of procrastinating I finally updated a set of FAQs that I compiled for colleagues back in 2023 – in part because I couldn't find anything better out there!
Back in February 2023, I sat down with ChatGPT and pulled together a set of frequently asked questions for college-level instructors on using the app.
Many of you will remember that this was a time when everything exploded around students using generative AI to “cheat” on assignments, and educators were running around like headless chickens trying to work out what to do!
At the time, many of my colleagues were struggling to find clear and to-the-point help and advice on ChatGPT. And so, in the absence of anything better, I decided to ask ChatGPT itself what instructors might need to know about its use in the classroom.
The result was a surprisingly useful set of frequently asked questions for anyone struggling to know where to start. But it quickly became outdated as things moved on, and I simply didn’t have the time to be constantly updating the FAQ.
This past week though, the college I’m a part of here at ASU issued some basic guidelines for faculty on artificial intelligence that linked to those original FAQ. And so, stung into action by the guilt of imagining colleagues using a long-outdated document, I set about updating them.
As in 2023, I turned to ChatGPT for help; it was a refreshing reminder of just how powerful the platform is when used effectively.
If you’re interested in the updated FAQ and want to cut to the chase, they’re linked to below as a Google Doc or PDF.
I did, however, want to say a little more about the process of updating them, as I think it highlights both the power of generative AI platforms like ChatGPT, and the surprising dearth of suitable guides for educators who are wondering where on earth to begin with what, to them, is still a new technology and uncharted territory.
The Process
One of the reasons I hesitated so long before updating the FAQ was the thought of how much work it would take — in the original version there are 32 questions, and I simply didn’t have the time to do the heavy lifting and research necessary to revise and update them all.
But, of course, this was me in pre-ChatGPT thinking mode.
When the penny dropped, I opened up ChatGPT (using version 4o), told it what I wanted to do, and started uploading old content and asking for updated text.
The process ended up being both iterative and generative. I started by uploading each or the original FAQ one by one. This quickly progressed to uploading the list of questions in each section and asking for ideas on revisions and updates before diving into specifics.
After we’d crafted a first updated draft of the FAQ together,1 I asked ChatGPT to review the whole question set and suggest additions and changes — with me selecting the ones that made sense. We then started refining the whole set of FAQ, shortening and tightening the language where necessary while adding new and updated content.
We finally went back and forth on further tweaks that I wanted to make to the revised FAQ — and had a discussion on what was still missing and how the questions could be updated to address this.
Through all of this, ChatGPT did the bulk of the work. My primary role was to curate everything and edited it where necessary to make sure that what we ended up made sense. (I also asked some questions that that helped the process along!)2
The result is most likely not the most polished or comprehensive set of FAQ for using text-based generative AI in the classroom.3 But it does ask the types of questions that educators who are floundering are likely to ask. And it provides clear and direct answers that I think present information that’s useful without getting lost in the weeds.
The FAQ are also much better than anything I could have achieved on my own — and substantially better than I suspect a committee could have produced.
And this really stuck out to me, having gone through the revisions.
This update took around 4 hours from start to finish. And as I was doing it I was also juggling meetings, interviews, and other stuff that was on my agenda for the day. So what you see is the result of around one and a half to two hours work with ChatGPT.
In contrast, if this had been produced by a university committee convened specifically for the task of creating a set of frequently asked questions on using ChatGPT in the classroom, my guess is that it would have taken 2 - 4 weeks to convene a committee of 5 - 10 people, which would then have spent between 3 - 6 months producing a mediocre set of FAQ.4
In effect, the comparison is between 4 hours’ work and a useful document using ChatGPT, and something like six months work representing over 200 person-hours that led to a document that might or might not be fit for purpose.
And this is perhaps my biggest takeaway from the process. Generative AI is limited and flawed. But it’s also a game changer in a growing number of situations when used appropriately. And a failure to realize or acknowledge this — especially in learning and education — is likely to lead to a growing gap between those who are succeeding because they’re learning how to make good use of emerging capabilities, and those who are failing because they’re not.
Again, the FAQ can be accessed using the links above — and are provided under a Creative Commons license so they can be freely shared and adapted (as long as this isn’t for commercial use).
Observant readers will notice the use of “we” here. I went back and forth on this, but in the end decided that, as this very much felt like a collaboration between ChatGPT and myself, the “we” was appropriate. I suspect some will disagree, but I’m not sure we have the appropriate language yet to describe working with machines that are fundamentally different from passive devices.
This process — as any regular users of generative AI will be familiar with — was one of ChatGPT substantially amplifying and extending my existing expertise. Without me having a good grasp of what I wanted and how to ensure it was fit for purpose, the FAQ would not have been as fit-for-purpose as they are. But working with ChatGPT both reduced the barrier to starting work on the FAQ, and massively reduced the time and effort needed to produce them.
I’m sure there are similar resources out there somewhere. But every time I look all I come across is convoluted guides that don’t tell me what I need to know, training courses that involve watching interminable videos or completing equally interminable assignments (OK, I may have been guilty of being part of this), or opinion pieces that focus on what the author wants to say, not what I want to read. If you know of anything comparable, please let me know!
I may be being a tad on the disingenuous side here — and apologies if you’re part of a fast and efficient academic committee that’s produced useful guidelines far faster than my estimate. On the other hand, I’m waiting for the emails telling me that my estimate was way too optimistic!
Thanks for doing this. Wondering if making this into a GPT, Claude Artifact, etc. would allow for a more "easy" way of keeping it current. Of course by making it CC I could do that myself.