Generative AI as sustainability futuring assistant

Teams of Masters students collaborate with GenAI to build rich pictures of sustainable future city precincts

Education category: Interdisciplinary unit (STEM+HASS, combination of design, social science, ecology and engineering concepts).

Assessment types applicable: Team precinct design challenge, delivered as 20 minute verbal/visual presentation.

Unit Context: In the Masters elective, ENS5340: Transforming Cities for Sustainability, students engage in a speculative, creative futuring process to envision and present a sustainable future for the Monash precinct in 2040. They also come up with innovative, specific projects we could take in-the-now to help desirable aspects of that future come to be. GenAI - text and image - tools were introduced in Jan 2023 as a creative task collaborator to broaden perspectives, generate rich future stories, visualise artefacts and scenes from student-created scenarios, and to suggest creative ways to communicate their future via a group presentation. Students could take forward any, or none, of the outputs into their graded presentation.

Learning Objectives: Unit Learning Outcome: Integrate disciplinary and paradigmatic approaches to design or redesign a case study urban location, grounded in theory, evidence and practice. Broader learning outcomes: responsibly and creatively engage with GenAI.

Which AI tools & why: ChatGPT/Bard/CoPilot; Midjourney; Tensor; DALL-E 2. 

How is task & AI structured/scaffolded: Students were provided a guide (linked here) which was introduced and tested in-class. Educators explained when AI was likely to be more or less useful in the creative and critical process, and in-class reflection on the ethics and utility of AI was carried out mid-way.

What support/guidance was needed for students: Workshop aid was provided above. In 2023, most students had not used these tools, so in-class support and testing was required; in 2024, most had used the tools, so more tailored content-driven guidance, or comparisons between tools, was prioritised.

How were responsible use and ethical issues addressed: Collective class reflection and debrief, + as part of supplied guidance material.

2-3 things that worked well: Diversity of tool options spoke to different learners - some loved the visuals and rapidly explored these, others used the text tools heavily. Social context was important - learners often ‘showing and telling’ each other live and sitting around a single laptop, with the AI tools stimulating creative discussion. Visual tools allowed students without graphic design/illustration skills to visualise future city projects (somewhat) more easily than otherwise could with most teams using outputs in presentation. Students took tools in unexpected directions such as exploring first-person narratives of disaster scenarios, writing poetry of sustainable futures from non-human perspectives, and even folding the AI tools into their sustainable scenarios as key features in the future of Monash precinct.

2-3 things that need improvement: Big gaps exist with student familiarity and confidence using these tools - some treat them as a glorified search engine, others are adept with many uses. A strong existing teamwork dynamic was good for this as learners helped each other learn, but if this is not present, learning gaps will persist.

Things to consider when adapting to your context: The tools were introduced in the context of a creative process, not just at the start, so teams had a clear purpose in mind when engaging with them. Thus, the tools expanded - not shortcut - their overall efforts. Given the open-ended and speculative task, ‘truth’/fidelity was not that important, so we needed less guardrails on how to use AI content ethically or with academic integrity.

Previous
Previous

Design simulator

Next
Next

Generative AI as an ideation assistant