I recently asked a group of around 3000 educators from my mailing list what they’d like to learn about GenAI. I’m incredibly lucky to have feet in a couple of camps as both an early career researcher and a consultant/author. It means I have access to the training and skills for working with lots of information, and a large audience to draw on when I’m looking for ideas. So, when I asked the question What specific topics or concepts related to AI would you most like to see? I knew I’d get plenty of interesting responses.
But this time, I was overwhelmed by responses. We all get survey fatigue, and blasting a semi-anonymous mailing list isn’t always an effective way of gathering feedback. This time, however, I received hundreds of detailed, engaged, and often incredibly passionate responses to the survey. In this post, I’m going to talk about some of the main threads and connect ideas across sectors and countries.
Educators all over the world are looking for advice
First of all, a little note on the people who took the time to respond to the survey. This information, which of course I’m anonymising here, came from the optional question“Tell me a little about yourself”. In fact, though it was optional, the vast majority of respondents answered this and it was often the longest response.
Every sector of education was represented: K-12; early childhood education; independent, faith, and government schools; mental health services; universities; TAFEs and vocational education; adult education; and industries related to education including IT, edtech, and administration.
Among these broad demographics were responses from Australia (Perth, Melbourne, Tassie, Sydney, SA and QLD amongst them), the US (including Illinois, New Hampshire, Oregon, California, and Florida), Argentina, South Africa, Denmark, the Netherlands) and the UK and Europe.
Some even volunteered more specifics, including years teaching (ranging from new graduates and students through to over 40 years’ experience) and subject disciplines (Chemistry, English, History, Geography, VET, Music, Arts, Science, Programming and IT, Software Development, Literature, Classics, and a whole range of varied University courses).
Before I get into the main question, I just wanted to shout out a huge thank you to everyone who took the time to respond! I know how hard it is to find time to even read emails, let alone respond to them, and especially for educators. Every single response was carefully read and appreciated.
What do educators want to learn?
So, what do educators actually want to know about this topic? To frame the question, it’s worth pointing out that the recipients of the survey are on a sub-group of my mailing list that has been receiving resources throughout July-August, including ebooks like Rethinking Assessment and Teaching AI Ethics. No doubt, these resources will have influenced the responses. One respondent even wrote a glowing review, which I’ll include here because my ego is as fragile as anyone else’s:
Thanks for the resources. It is extremely rare to find someone willing to share such a substantial amount of information for free. I think you have done a wonderful job.
You’re welcome 😉
If you’re interested in those resources they’re all available for free until the end of September, and the only trade-off is that I’ll email you surveys asking questions like these.
On with the show…
How can students use AI without losing their critical and creative thinking skills?
By far one of the most common questions from respondents was along the lines of How can students use AI without losing their critical and creative thinking skills? The actual language varied, including, “how do we support students to use AI without it taking all their ideas away?” and “how can students use AI without compromising their education?”
Personally, I believe critical and creative thinking skills are innate – they can of course be developed, but every child I have ever met is curious and interested (even if what they’re interested in is not necessarily what I’m teaching…). AI is only a threat to those talents if we buy into the tech company line of “inevitability”, which I don’t subscribe to. But I do think that, used carelessly or ignored, AI is going to potentially do more harm than good in the classroom.
I often joke – and it’s only half a joke – that I could teach 90% of my Literature class without a computer, let alone AI. But you can be sure that the remaining 10% would feature some solid instruction on AI, digital texts, and what technology means for writing and the Arts. It’s about finding that relevance and balance. Not just pouring AI special sauce all over the curriculum for the sake of it, but consciously and critically discussing the technology when and where it’s appropriate.
How can we learn (and teach) about AI ethics?
The ethical concerns of AI were a prominent feature throughout the responses, both in terms of teacher and student use. Respondents asked about bias, the environment, and in one excellently crafted question the growing digital divide represented by the gap between the best (most expensive) AI systems and those available for free.
Teaching AI Ethics is important to me, so much so that I’ve written about it extensively on this blog and published an open access ebook of articles and teaching ideas. It’s obviously important to other educators too, as over half of the responses featured one or more requests for topics related to AI ethics.
How might we use AI to create high quality resources?
Educators are sick of crappy lesson plan generating buttons produced by companies like Google who, frankly, should know better by now. Some of these kinds of response – asking how educators might use the technology well – bordered on angry. Take this one for example:
I’m sick of tech companies telling me how to teach, and now they’re telling me how to make the resources I teach WITH. I’m not interested in autogenerated lesson plans and worksheets. I haven’t used a worksheet since I was a graduate teacher! I’m not going back. But I have used chatGPT to make useful resources based on the [local] curriculum and I’d love to know more about how to best use AI in ways which support but don’t replace all of my hard-earned expertise.
This sentiment was echoes again and again, with educators asking for tips on using AI, and not being used by it.
How do we rethink assessment?
This was, of course, another of the most commonly asked questions. Variations included thoughts on redesigning existing assessments, whether we need to throw everything out entirely, and how to convince colleagues (and entire systems) that change is necessary.
Just prior to reviewing the responses, an interesting article was published by Thomas Corbin, Margaret Bearman, David Boud, and Phillip Dawson, titled The wicked problem of AI and assessment. Go and read it yourself, but in a nutshell the paper argues that GenAI, “resists definitive formulation, offers only better or worse rather than correct solutions, cannot be tested without consequence, and places significant responsibility on decision-makers.”
A lot of this came through in the responses to my survey from across education sectors and countries. Some had attempted to integrate our AI Assessment Scale, and were hitting various walls at the individual and system levels. Others were exploring frameworks like University of Sydney’s ‘Two Lanes’ approach, or adaptations of scales and frameworks from around the world. Everyone is facing challenges and looking for advice.
One thing we’re all clear on is that things can’t stay the same.
Where do we find the time?
And last of all (though there were many questions that didn’t fall neatly into these five categories), where do educators find the time to learn all of this? As I argued recently in a post about professional development, teachers have to develop technological, domain, and situated expertise. That means not just understanding AI, but also understanding their subject or discipline, and developing the contextual, lived experience of a skilled teacher over time.
All of this takes a tremendous amount of time and effort, and teachers are notoriously time-poor. So where do we find time for the additional skills of GenAI? And with the technology developing so quickly, how can we keep up?
A second question on my survey asked “Do you have any suggestions regarding the structure, pacing, or any other aspects of a new AI course?” I have a few ideas of my own based on my experiences as a teacher, but it’s always good to hear back from the community how they’d like to tackle PD. Plenty of research (including my own PhD) backs up the idea that a huge part of teacher PD, and especially digital, happens in our own time. Whether that’s appropriate or not (it isn’t) is not the question here: it’s a fact.
To learn about GenAI, the respondents indicated they’d like online, asynchronous, bit-sized PD sessions. Courses which are self-paced but have a clear schedule for completion and some kind of accountability process. Not too long, but not so short that it’s forgettable. Actionable in and out of the classroom, and something which could be done during office hours if possible, such as during breaks or planning time.
New courses coming soon
As I mentioned earlier, the reason I’m gathering all of this information – other than my own personal interest – is because I’m working on my 2025 courses.
Currently, over 3000 educators are enrolled on courses on my platform at Practical AI Strategies, including in the 2024 4.5 hour flagship course of the same name. Since this technology develops constantly, some of that material is due for a refresh. But rather than just updating the old courses, I’m making some entirely new ones.
The free resources which have been sent out in the past months were a way of saying thank you to all of the educators who have joined me in the past few years in trying to understand the complexities of GenAI.
And thanks again to everyone who’s responded so far!
Leave a Reply