It’s been brought to my attention that my use of AI imagery might be getting out of control. So I am using this as an opportunity to try and make BetterU look more like the community I want it to serve.
I recently attended a meeting of Student Council to talk about BetterU, the student wellbeing portal and hub. Thank you to the council for finding me a space on the agenda 🙏🏼
One of the things I spoke about was the shift we are trying to make towards more student-created content.
The Student Health and Wellbeing Blog, which is the article library that preceded BetterU, has been running since 2017. Across the life of the blog, only around 3% of posts were student-created. Over the past 12 months however, we’ve been trying to change that, and it appears we are succeeding. At the moment, we are around 30% student-created content, with the longer-term goal of getting closer to 50%.
That shift has been happening in a few different ways: student placements with BetterU, including social work, public health and psychology; invited contributions from students doing relevant wellbeing-related work; engagement with Oasis wellbeing ambassadors; invitations to students whose work we have noticed elsewhere, such as in Empire Times; and a more general call-out for students interested in contributing.
The goal is for BetterU to become less of a service talking to students, and more of a community magazine shaped with students.
At the Student Council meeting, I also asked for broader comments or reactions to BetterU.
One of the students raised the amount of AI-generated imagery on the site and asked why we used so much of it.
My answer was that as far as managing the site goes, it is mostly just me. I don’t have significant creative or design skills, and a lot of posts have short turnaround times. AI imagery has been useful because it can quickly create images that match the themes of posts more closely than generic stock imagery.
But the student then said something that stuck with me: AI imagery can be a turn-off. It can make them less likely to visit a site or read an article. And then a second student agreed. I’ve been pondering on that feedback since then.
Then I was talking with Ben, the manager of the Flinders University Student Association, and I mentioned the conversation. He said FUSA had received similar feedback. Their media team had responded by trying to make imagery as hyper-local as possible: Flinders people, Flinders places, Flinders events, Flinders life.
So I have a new project: to start changing the imagery landscape on BetterU.
I can’t promise that AI imagery will disappear entirely. There are times when I think it can work really well. For example, Jingyi (Lucy), who is currently on placement with us from Public Health, recently wrote a great piece on productivity guilt. She used AI to generate a set of images that told different parts of the story. The images were thematically connected to the article and to each other, and I think they worked nicely. When AI imagery is used thoughtfully and deliberately, it can do a good job.
But I don’t think that for the most part I have been using it that way. Often I use it because it is quick and easy. This is the kind of use that I want to reduce and that is on me.
So what is the goal?
I want to increase the amount of Flinders community-generated imagery on BetterU. I set the goal of 50% for student-created content. I think a goal of 75% community generated imagery is possible. And I think there are practical and achievable ways to do it.
Here are some of the things I am going to explore.
First, I want to connect with people in media and communications across Flinders to see whether there are existing repositories of Flinders-specific images that could be used more often. I actually have a meeting this afternoon with the comms engagement group, so the timing on this one is excellent.
Second, I want to make imagery generation — photography, illustration, design, visual storytelling — more of a focus for students on placement with BetterU. At the moment, I have students focus most of their attention on the written content, the narrative. That is because they are students in health and welfare disciplines and the goal of their placements is to become good communicators of health and wellbeing information. But it is clearer to me now that we can’t separate out the imagery we use from the health messages we are sending.
Third, I want to invite students and staff (that is you, reader) to submit media they have created. That might be campus photography, drawings, paintings, digital art, or other forms of visual imagery. With appropriate permission and attribution, we could then catalogue that material and use it in future posts where there is a good match between the image and the content.
Fourth, I want to learn more myself about how to generate novel and interesting imagery for the site. I have a lot of respect for the work done by student media, including Steph and the team at Empire Times, who continue to produce engaging, distinctive and creative visuals. I’d like to learn from that kind of approach.
Fifth, I want to look for topics, courses or student projects where imagery generation is part of the degree, and see whether there are possible collaborations.
In short: I am committing to reducing the amount of quickly generated AI imagery on BetterU and increasing the amount of Flinders community-generated imagery.
So, if you are still reading, consider this an informal invitation.
If you are a Flinders student or staff member and you have campus photography, artwork, illustration, design work, or other visual imagery that you would be interested in sharing and potentially having featured in future BetterU posts, please get in contact with me. You can email me at gareth.furber@flinders.edu.au.
I’d love to start building a catalogue of images that can help BetterU look more like the community it represents.
And finally, I want to say thank you to the two students at the Student Council meeting who raised this issue. I heard you and I’m going to try to do something useful with it.
