"Big tech AI normalises unsustainable consumption – but no one takes responsibility for the emissions"
Search engines and online marketplaces like Google and Amazon are shaping how we live, shop and travel. In doing so, they normalise carbon-intensive lifestyles – but the emissions that follows are not counted as theirs. Researchers within Mistra Environmental Communication argue that this must change.
Jutta Haider
Photo: Julieta Longo & Digit / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/
Search engines are an integrated but often invisible part of how we navigate the world. They shape our understanding of reality and can thereby harm the environment. How, you may ask?
The emissions from servers, cooling systems and office buildings are only a part of the total amount that big tech companies emit. These companies also facilitate and normalise high-emission lifestyles and consumption practices.
Let us give you an example of how. When you shop online, new products are prioritised over second-hand or repair by default. And when you search for travel alternatives, flight planners are higher ranked than more sustainable public transportation. The algorithms are steering us towards high-emission choices.
AI strengthens these consumption biases even more. However, since AI-driven interactions are so common, the algorithmic decisions behind them usually go unnoticed. This results in consumption-based emissions that are never attributed to the tech giants. But they should be. We propose to call them algorithmically facilitated emissions – algorithmic decisions that prioritise high-carbon practices and lifestyles.
As the climate crisis accelerates, it is high time to start including algorithmically facilitated emissions in regulatory frameworks and in companies’ emission reporting. By recognising this issue, the responsibility for this kind of climate impact can be placed at the source.
This post was written by Jutta Haider, Professor in information studies and Leader of the focus area Information, with support by communications officer Clara Jonsson.
"Big tech AI normalises unsustainable consumption – but no one takes responsibility for the emissions"
Search engines are an integrated but often invisible part of how we navigate the world. They shape our understanding of reality and can thereby harm the environment. How, you may ask?
The emissions from servers, cooling systems and office buildings are only a part of the total amount that big tech companies emit. These companies also facilitate and normalise high-emission lifestyles and consumption practices.
Let us give you an example of how. When you shop online, new products are prioritised over second-hand or repair by default. And when you search for travel alternatives, flight planners are higher ranked than more sustainable public transportation. The algorithms are steering us towards high-emission choices.
AI strengthens these consumption biases even more. However, since AI-driven interactions are so common, the algorithmic decisions behind them usually go unnoticed. This results in consumption-based emissions that are never attributed to the tech giants. But they should be. We propose to call them algorithmically facilitated emissions – algorithmic decisions that prioritise high-carbon practices and lifestyles.
As the climate crisis accelerates, it is high time to start including algorithmically facilitated emissions in regulatory frameworks and in companies’ emission reporting. By recognising this issue, the responsibility for this kind of climate impact can be placed at the source.
Read the full commentary: Unsustainable artificial intelligence and algorithmically facilitated emissions: The case for emissions-reduction-by-design
This post was written by Jutta Haider, Professor in information studies and Leader of the focus area Information, with support by communications officer Clara Jonsson.