We have been working with a group of around twenty Grade 8 students on the theme of futures thinking – with a specific focus on anticipating future forms of schooling in an age of AI and climate change.

This is a fascinating – but underdiscussed – topic that has only recently begun to be talked about research and policy circles … and certainly not something that features prominently in mainstream discussions about the future of education.

On day two, the students formed small groups and were led through prolonged bouts of scenario-building using a series of prompts, provocations and methods. They came up with insightful, innovative and imaginative takes on ‘probable’, ‘possible’, ‘plausible’ and ‘preferable’ school futures – all infused by various anticipated developments in AI, energy and environmental developments.

Then, on day three, the five groups were tasked with producing short presentations about their scenarios. Each group was given free rein on how to approach this. They could use any analogue and digital materials to hand. They were welcome to produce PowerPoint pitches,  but were encouraged to perhaps choose to produce an artwork or act out a dramatization to illustrate their anticipated future … anything that grabbed their imagination.

As it turned out, all groups opted to give standard laptop presentations using the classroom overhead projector and taking turns to narrate segments of prepared scripts. The students were clearly proficient in this style of presentation and used an impressive array of apps and software.

Yet, while these were all highly competent presentations there was a pronounced flattening-out of the ‘future school’ ideas and arguments that we had heard on Day 2. All of the presentations looked, read and sounded like slick corporate pitches, but their content tended to fall short of the interesting, quirky, speculative features of the Day 2 conversations. 

As the presentations progressed I began to wonder if this was a side-effect of the students’ use of online sources and software to prepare and present their ideas. Certainly,  a lot of the rough edges and interesting wrinkles felt like they had been smoothed out in the translation from workshop brainstorming to the PowerPoint slide deck.

For example, in visual terms these were all very professional-looking presentations. One group has used Canva to produce a slide deck that would have passed easily for a tech-firm boardroom pitch to shareholders. This featured sleek graphics of robot heads, humans wearing headsets, and similar visual tropes associated with ‘future’ and ‘AI’ keywords. Another group had used Wix to present their scenario in website form – clicking through cleanly designed pages labelled About, School Services, Timing and Contact Us.

Lots of these presentations featured graphics of ‘future schools’ – almost always CGI visions of sleek curved buildings, vast atriums and library spaces, rather than any other non-architectural aspect of ‘school’.  The group that had created images using Dall:E featured similarly generic images of multi-storey buildings with curved glass-fronted facades, living walls and other architectural tropes from the late 2010s.

Another group had reappropriated a corporate video of the ‘university of future’ from a glass company who were clearly looking to sell to higher education customers. The students had repurposed the company’s computer-generated animations of glass-fronted rooms, glass geo-domes and glass-writing pads – sneakily blurring out the company name and overlaying their own text about their future school. 

All of this resulted in presentations that undoubtedly looked fantastic, but felt completely in lockstep with the endless corporate ‘socio-technical imaginaries’ of schools and education that we see from the likes of KPMG, OECD, and numerous other sources.

This parroting of familiar ‘future school’ tropes was even more evident in most of the students’ accompanying texts.  Notably, many of the groups had made use of Chat GPT to regurgitate appropriately vague descriptions of future schools. We saw a succession of slides pushing familiar-sounding promises of “cutting-edge educational spaces that seamlessly integrate advanced technology and innovative design” and “a commitment to knowledge, innovation and environmental responsibility”.

The full effect of this generic future-speak is illustrated by this group slide:

“Gotham Sigma School in the year 2070 boasts cutting-edge futuristic facilities that redefine the educational landscape. The architecture is characterised by sustainable energy-efficient structures with transparent surfaces, and holographic displays welcome students with real-time information. Classrooms feature immersive VR and AR technology, smart desks, and adaptive learning interfaces. The library is a multimedia hub with holographic books displays, while sports facilities include virtual simulation and advanced training rooms … The school prioritises sustainability with green spaces incorporating vertical gardens … Security measures include biometrics, AI-driven surveillance and emergency protocols. The school epitomises a harmonious blend of technology, sustainability and innovation, fostering holistic development in students”

Tellingly, very little of the more imaginative, left-field written descriptions that had been discussed during the Day 2 scenario-building workshops made it through to these slide-decks. 

One student had built a simple chat-bot to handle his group’s Q&A. This was presented as the ‘School Owner’, and the group insisted on passing all of the audience questions through the bot, which reliably spat back very assured answers such as:

“Yes, this school is a co-ed school welcoming students and staff from all genders. We believe in equal opportunity for everyone and promote inclusive and diverse learning environments, where everyone is welcome”.

This chatbot had been built using C.AI, which the student told us could be trained on as little as 30,000 words of text. The group had produced this 30,000 words of text on their imagined school using Chat GPT. The ‘School Owner’ was therefore artificially generating short text responses based on this larger body of artificially generated text … resulting in bland recursive ‘answers’ that appeared relevant but were oddly empty of any real spirit.

Another group started by presenting their school logo which featured an intriguing motto which I thought might perhaps have some hidden meaning (‘Somama Soth Griam’) … but this turned out simply to be text that was randomly generated by the Dall:E generative AI image-making tool.

That said, these digitally-flattened presentations did show occasional flashes of imagination. Intriguingly, three of the different group presentations contained relatively innovative ideas relating to school food. For example, one group suggested AI in the canteen that could conduct health analyses of each new diner and prepare the most appropriate meal for their health needs. Another canteen was based around 3D printed food. Also relating to food, another group included a throwaway line about their future school having a five-star Michelin restaurant. 

On reflection, these might well reflect the tropes and thinking circulating online regarding food futures. It might well be that typing in prompts in Chat GPT and Google Bard about “future canteens” or “future food” comes up with more interesting responses than prompts asking for “future school”.

Otherwise, there was very little to suggest that these ideas had not come from a team of consultants or planners … rather than a group of fourteen year-olds from the Melbourne suburbs.

All of this brought home the ease with which corporate visions of the future school are replicated and regurgitated through online Google image searches and generative AI prompts. If the main source material on the internet relating to ‘future schools’ is as limited as it plainly is, then any attempt by imaginative 14 years old’s to digitally research and present ideas of ‘future school’ is going to echo, mirror and replicate these corporate imaginaries.

Interestingly, the one group presentation that stood out was by a group of boys who had clearly not taken much care and time with their presentation. While their presentation was no-where near as slick and coherent than their others, in some ways their scenario seemed the most honest (and perhaps more human). This group’s text certainly seemed more likely to have been produced in a hurry by a group of by school students born in 2009. Take this sample quote from their opening slide: 

“As years have passed, technology has revolutionised in many ways. For example back in 80’s and 90’s people did not know what a phone was even, or not one had refrigerators, laptops and etc. However, the word technology has come to the rescue and has changed our world in many different ways … We as a group think schools will have way more technology, and the classrooms will be made out of different materials as supposed to currently being made out of bricks and wood”.

In one sense, this presentation was not nearly as comprehensive or coherent than the earlier pitches. Yet it did contain more idiosyncratic thinking that prompted far more questions and discussion than the other groups. For example, this group proposed a global chain of over 100 schools to accommodate increased levels of migration around the world. This led to questions from the audience about how each school could fulfil the group’s aim of tuition in the local languages. It also led to a pointed question from another students why there were no schools located on the African continent … to which the group responded that it ‘would be too difficult’. This more messy exchange stood in sharp contrast to the earlier chat-bot response that “We believe in equal opportunity for everyone”. 

All told, this group presentation exercise highlighted the constraining and homogenising effect of thinking about the future of education through the medium of PowerPoint, Canva, generative AI and web-searches for existing ‘future schools’ content. The use of these online tools to illustrate the students’ initial ideas resulted in the regurgitation of ‘future’ tropes around education. In contrast to the Day 2 activities that we led by butcher’s paper, sticky notes, provocation cards and group discussions, the Day 3 presentations turned out to be a frustratingly restrictive – rather than expansive – exercise.

This all leads to a rather old-fashioned conclusion … if we want young people to think expansively and differently about the future of schooling in an age of emerging technology, perhaps there is value in encouraging them to stay offline for as long as possible?