This post presents the main findings of the seed funding project Environmental Sustainability in MishMash, led by Lina Plataniti (NORSUS) in collaboration with SINTEF Digital. Lina will present the work at poster 38 during the MishMash opening conference in Kristiansand on 14–15 September.

Sustainable use of AI in MishMash: the beginning

NORSUS invited SINTEF Digital to join a seed funding project in MishMash to explore the environmental sustainability of AI in the context of art, creativity and culture. In June 2026, we had a workshop in Oslo with the goal to initiate the dialogue and work on AI and environmental sustainability in MishMash.

Why does sustainability matter?

“Our Common Future”, the report of the World Commission on Environment and Development¹, which, it is worth mentioning, was chaired by Gro Harlem Brundtland, Norway’s first female prime minister, gave this classic definition of sustainable development:

“the development that meets the needs of the present without compromising the ability of future generations to meet their own needs.”

Human activities are pushing the Earth system beyond the ideal, stable conditions under which humanity has thrived, making sustainable development sound rather utopian nowadays. Seven out of nine Planetary Boundaries² have already been transgressed according to the latest scientific updates³. It is important to note that by referring to the “Planet” we should not undermine the regional or local effects of these conditions, which extend beyond the natural environment to communities and human well-being.

The contribution of information technology to humanity is extraordinary, however, it does have impacts that may intensify the transgression of Planetary Boundaries. AI is a groundbreaking technological development, but is it sustainable? And what might “sustainable use of AI” mean for MishMash?

The workshop

We are a team of sustainability scientists and green AI / IT researchers who joined forces in a 3-hour workshop to explore and map what environmentally sustainable use of AI could imply for MishMash. While the initial intention was to look only into the environmental impacts of AI, early in our discussions we realised that we preferred not to exclude the potential social or socio-ethical implications. Nevertheless, the environmental aspect remains central to our work.

Within this scope we attempted to put on paper what sustainability may imply for MishMash through three exercises:

1. definition of environmental sustainability in MishMash,
2. mapping of our research interests and competencies, and
3. planning future projects.

Sustainability in MishMash

In our attempt to find out how to approach the topic and define sustainability in MishMash, we identified the following key points:

• It is important to raise awareness about the environmental impacts of AI. The users should become aware of what lies behind the interfaces they interact with when using AI.
• AI relies on a complex system of machines and networks that is resource-intensive both for constructing and operating it. Large amounts of energy, scarce and critical raw materials, land use transformation, and extensive water consumption support the value chain behind the technological boom of AI.
• It is usually easier to describe than measure these impacts, due to a lack of transparent data from the providers of AI technologies.
• The environmental impacts described above should not be seen separately from the social implications stemming from climate change (mostly related to energy use), extraction of minerals (e.g. local air, soil and water pollution, working conditions in mines, etc.), competition of access to common goods like water and land, heat-islands and noise pollution around data centres.

We found it necessary to understand the perspective of the AI users: why and to what extent they use AI in artistic, creative or cultural work. Are they aiming for better art or more art? The importance of involving other interested MishMash members in an interdisciplinary dialogue through similar workshops, interviews, and common projects became clear.

Research interests

We would like to adopt a holistic approach when studying the impacts of AI systems, considering them as tangible products with specific lifecycles: from raw material extraction and manufacturing to use and end-of-life. Life Cycle Assessment (LCA) is a well-established methodology aligned with this approach and commonly used by sustainability scientists. In practice, this is rather complex and presents methodological challenges.

Methodologically it is important to define the scope of the assessment as well as the AI system itself. Are we studying the hardware or the whole service? What is our reference, which function does our system represent? For example, is it the traffic of 1M tokens, the creation of one song, or the enjoyment that an AI art piece creates to the audience? At the same time multiple lifecycles coexist in one AI system: the lifecycles of the material components and infrastructure (chips, servers, data centres, networks, end-user devices) and the lifecycle of the software itself (data and energy used during development, training, use, deprecation of AI models). Another challenge is the creation of the Life Cycle Inventory, the dataset that provides information about all the components of the studied system, e.g. physical components, energy use or data centre location.

A useful exercise with LCA would be to work with specific case studies comparing the impacts of artistic practices with and without the use of AI, for example in film making, songwriting, musical or theatrical productions.

Another very appealing research approach is to adopt and implement sustainability principles in AI models used in creative work. Can we extend the lifetime of an AI model? Can we repurpose an old AI model? Can we reuse or recycle deprecated AI models? Can we do frugal computing? Our IT researchers claimed that such concepts are possible, a prospect that was met with great enthusiasm by the sustainability scientists!

Further, we would like to investigate how the AI infrastructure and the environmental impacts affect social and cultural activities at local communities. AI can be beneficial for the preservation of cultural heritage through archives, for example. Do such benefits exceed the costs of e.g. the co-existence of data centres and indigenous areas in Northern Norway? In such cases a sound methodological approach would be to combine environmental LCA with qualitative social LCA.

All in all, our disciplines appear to be complementing each other, opening the way for interesting collaborations!

How could we contribute to MishMash?

Our contributions to MishMash can be manifold. As a first step, we need to get better understanding of the artistic and creative side of the coin and enrich our dialogue with contributions from artists, creators, curators, museologists, archivists, educators, and others. Here comes a list with our main ideas:

• Involve MishMash members from other disciplines through direct dialogue (interviews, meetings) to understand their perspectives and needs, and explore project collaborations.
• Conduct hands-on work: life cycle assessments (LCAs) in case-studies. Examples are films, performance productions, song writing, and other. Please contact us if you would like to collaborate on such projects.
• Create a feedback tool that provides information about the impacts of AI tools and recommendations for reducing them, with the goal to support the users’ decision-making regarding the most “sustainable” paths to work with AI.
• Use artistic and creative approaches to present our findings and messages, always in a useful and easy to grasp way.

Next step: our survey

As mentioned, we would like to invite contributions from all interested MishMash members. We have created a survey to find out more about creators’ perspectives regarding environmental sustainability and sustainable use of AI in their work. Members who wish to participate in future collaborations are also encouraged to express their interest. The survey link will be shared through the MishMash mailing lists (subscribe to the announcements list if you are not on it yet).

For more information, please contact Lina Plataniti (lina@norsus.no).

1. World Commission on Environment and Development (WCED) (1987). Our Common Future. Oxford: Oxford University Press. ↩

2. Potsdam Institute for Climate Impact Research – FAQ. The Planetary Boundaries are a scientific framework that defines “nine biophysical systems and processes that regulate the functioning of life support systems on Earth and ultimately the stability and resilience of the Earth system”. ↩

3. Planetary Health Check ↩