Generating Images on the Urgency of Climate Responsibility

This paper explores how art created through the application of Artificial Intelligence (AI) systems can go beyond a visual representation of the climate crisis to instead engage in meaningful debate on the urgency of our climate responsibility and seek potential solutions.

The Arctic Circle is ground zero of climate change. The Arctic Ocean’s ice cover, which helps determine the Earth’s climate, fell to its second lowest level on record as of 7th September 2020. Humanity is dependent on the ocean and cryosphere. It interconnects with the climate system through water, energy, and carbon. The impact of this melting ice cover is also political, military, and economic as several nations vie for ownership and control over its greater navigable waters – a new Northwest Passage – and the opportunities it presents.

In April 2022, artist Pat Naldi, undertook a three-week research expedition to the high Arctic Archipelago aboard a Barquentine sailing vessel, as part of The Arctic Circle Artist and Scientist Residency Program. Sailing and making landings along the Svalbard archipelago – which is warming at the fastest rate anywhere - she bore witness to the melting ice cover and receding glaciers of the Arctic. A selection of Naldi’s analog colour film photographs of melting icebergs and glaciers shot on the expedition with a vintage 102-year-old Box Brownie camera, have been programmed into an AI specially created by the artist Anamarija Podrebarac. Named Polar Bear after the hypercarnivorous inhabitant of the Arctic Circle that is on the front line of the climate crisis relying as it does on sea ice to hunt for food, this AI is analysing the information provided by the analogue images of this climate affected environment and generating digital images of the Arctic landscape.

The energy consumption of AI systems, specifically machine learning, has itself, come under scrutiny, yet despite this, AI systems have the potential to decouple economic growth from rising carbon emissions and environmental degradation. It can halt emissions in the energy sector by forecasting the supply and demand of power in the grid, improve the scheduling of renewables, and reducing the life-cycle of fossil fuel emissions through predictive maintenance. By harnessing the swaths of data from sensors and satellites, AI can also better predict climate change impacts and proactively steward these ecosystems. We can also actively increase the capacity of carbon sinks like peatlands and accelerate afforestation through locating appropriate planting sites, monitoring plant health, and even controlling tree-planting drones.

Therefore, what can we learn about climate change from a collaboration between vintage analog imagery and contemporary Artificial Intelligence? Can these data generated images by Polar Bear AI, offer us a future glimpse of the effect of melting ice cover on global ecosystems, and thereby seek active solutions?