** Harnessing AI for Sustainability**
The world stands at a pivotal moment, confronting a polycrisis: climate change, biodiversity loss, pollution, inequality, wars and more. Artificial Intelligence (AI) is advancing at an extremely high rate, emerging as a potentially powerful ally in our mission for a flourishing sustainable future.
This pre-study shows examples of how AI can be a unique tool that can help us take transformative steps that previously have been locked in. Based on insights from 35 experts across diverse fields, we examine several broad societal AI opportunities and specific impacts within the food retail sector.
The goal of the pre-study is to spark engagement around these AI opportunities and form collaborations that can take leadership in directing and funding AI development towards the applications needed for an accelerated green transformation. We invite you to join us in this endeavour and wish you an inspiring reading experience.
The current AI Horizon
For most of its 70-year history, AI has developed at a glacial pace, but today the speed of development is such that any statements about the nature, impact, or limitations of AI are sure to be dated within months, if not weeks.
Previously confined to academic research, today AI technologies have escaped the lab. They are applied broadly—from cereal crop farming to urban planning and pharmaceutical research—as well as deeply—from low-level industrial systems all the way up to chatbots and digital assistants.
We are currently in an AI wave which bears all the hallmarks of previous technology-driven paradigm shifts. From recent examples such as the mobile revolution or the internet boom, to older ones like the first age of electrification, these shifts have some common characteristics. They disrupt existing systems, trigger rapid changes in the economy, and alter human behaviour and wider societal norms. They are not evenly distributed, but their impact is always global and felt on a timescale of decades.
We do not know how AI will be applied even a single decade from now. Broadly we can say that AI is likely to grow as an interface between users and complexity. Some AI will continue to passively analyse, recognise, and predict, delivering results to users for action. In other applications AI will likely have agency, managing and controlling systems of varying complexity. AI may soon help you buy a train ticket as well as run the traffic control systems of the train network.
The overall impact is even harder to predict, but AI can become a net positive for the planet. Whether this happens depends on how and where we apply the technology. A sustainable future of applied AI requires the active participation of people who are focused on sustainability. People who develop and showcase applications for sustainability, highlight the positive impact of others, drive change and advocate for effective regulation where that is needed.
In the midst of a rapidly developing present we are certain of this: the future of AI will be determined by those who engage in its development and use.
Building on Expert Input
With the support of AI Sweden, we conducted hour-long interviews with 25 highly knowledgeable professionals to understand the most serious barriers to green transformation.
The interviewees were selected to ensure a diverse range of perspectives. We focused on two scopes: a general scope not limited to any sector, and a specific focus on the food retail sector to achieve more concrete and specific results.
Next, we consulted 10 AI experts to explore AI opportunities that could address some of the identified barriers.
The authors then chose, synthesised, made hypotheses and simplifications based on this input. Consequently, a lot of great input was omitted for the sake of creating an introductory pre-study with the primary aim of sparking further conversation.
What we leave out
AI, like any powerful tool, comes with significant risks and potential downsides.
The most concerning include its misuse for harmful purposes or losing control over it. AI also has limitations that can make it unreliable, biased, unethical and basically could make many things worse. None of these aspects are covered in this pre-study.
There are two main concepts: ”Green AI” and ”AI for Green”. This pre-study focuses on the latter—using AI to support green transformation. ”Green AI” refers to making AI itself sustainable and resource-efficient, which is crucial as some AI applications significantly increase energy, mineral, and water consumption. However, some AI applications, like edge computing, are very resource-efficient.
The authors acknowledge the cultural biases of the interview group and of us, which limit the universality of the views in this pre-study. It is primarily based on experiences from the Global North. Therefore, when we refer to “us,” our perspective is limited in many ways. We invite readers to enrich the discussion with insights from diverse identities and experiences.
** AI'S Unique Potential **
The following sections delve into some of the challenges where we are ”stuck” today. We want to show that AI can play a unique role in providing opportunities for progress that previously have been locked in.
We’ve organised our discussion around pairs of challenges and actionable opportunities, providing a straightforward approach.
We begin with a wider perspective on the individual and societal benefits that AI brings to the table in situations where the information reaching us is of unnecessarily low quality. We then narrow our focus to food retail where data often is abundant and of good quality—a sector ripe for transformation through AI intervention.
** Harnessing AI for Sustainability**
The world stands at a pivotal moment, confronting a polycrisis: climate change, biodiversity loss, pollution, inequality, wars and more. Artificial Intelligence (AI) is advancing at an extremely high rate, emerging as a potentially powerful ally in our mission for a flourishing sustainable future.
This pre-study shows examples of how AI can be a unique tool that can help us take transformative steps that previously have been locked in. Based on insights from 35 experts across diverse fields, we examine several broad societal AI opportunities and specific impacts within the food retail sector.
The goal of the pre-study is to spark engagement around these AI opportunities and form collaborations that can take leadership in directing and funding AI development towards the applications needed for an accelerated green transformation. We invite you to join us in this endeavour and wish you an inspiring reading experience.
The current AI Horizon
For most of its 70-year history, AI has developed at a glacial pace, but today the speed of development is such that any statements about the nature, impact, or limitations of AI are sure to be dated within months, if not weeks.
Previously confined to academic research, today AI technologies have escaped the lab. They are applied broadly—from cereal crop farming to urban planning and pharmaceutical research—as well as deeply—from low-level industrial systems all the way up to chatbots and digital assistants.
We are currently in an AI wave which bears all the hallmarks of previous technology-driven paradigm shifts. From recent examples such as the mobile revolution or the internet boom, to older ones like the first age of electrification, these shifts have some common characteristics. They disrupt existing systems, trigger rapid changes in the economy, and alter human behaviour and wider societal norms. They are not evenly distributed, but their impact is always global and felt on a timescale of decades.
We do not know how AI will be applied even a single decade from now. Broadly we can say that AI is likely to grow as an interface between users and complexity. Some AI will continue to passively analyse, recognise, and predict, delivering results to users for action. In other applications AI will likely have agency, managing and controlling systems of varying complexity. AI may soon help you buy a train ticket as well as run the traffic control systems of the train network.
The overall impact is even harder to predict, but AI can become a net positive for the planet. Whether this happens depends on how and where we apply the technology. A sustainable future of applied AI requires the active participation of people who are focused on sustainability. People who develop and showcase applications for sustainability, highlight the positive impact of others, drive change and advocate for effective regulation where that is needed.
In the midst of a rapidly developing present we are certain of this: the future of AI will be determined by those who engage in its development and use.
Building on Expert Input
With the support of AI Sweden, we conducted hour-long interviews with 25 highly knowledgeable professionals to understand the most serious barriers to green transformation.
The interviewees were selected to ensure a diverse range of perspectives. We focused on two scopes: a general scope not limited to any sector, and a specific focus on the food retail sector to achieve more concrete and specific results.
Next, we consulted 10 AI experts to explore AI opportunities that could address some of the identified barriers.
The authors then chose, synthesised, made hypotheses and simplifications based on this input. Consequently, a lot of great input was omitted for the sake of creating an introductory pre-study with the primary aim of sparking further conversation.
What we leave out
AI, like any powerful tool, comes with significant risks and potential downsides.
The most concerning include its misuse for harmful purposes or losing control over it. AI also has limitations that can make it unreliable, biased, unethical and basically could make many things worse. None of these aspects are covered in this pre-study.
There are two main concepts: ”Green AI” and ”AI for Green”. This pre-study focuses on the latter—using AI to support green transformation. ”Green AI” refers to making AI itself sustainable and resource-efficient, which is crucial as some AI applications significantly increase energy, mineral, and water consumption. However, some AI applications, like edge computing, are very resource-efficient.
The authors acknowledge the cultural biases of the interview group and of us, which limit the universality of the views in this pre-study. It is primarily based on experiences from the Global North. Therefore, when we refer to “us,” our perspective is limited in many ways. We invite readers to enrich the discussion with insights from diverse identities and experiences.
** AI'S Unique Potential **
The following sections delve into some of the challenges where we are ”stuck” today. We want to show that AI can play a unique role in providing opportunities for progress that previously have been locked in.
We’ve organised our discussion around pairs of challenges and actionable opportunities, providing a straightforward approach.
We begin with a wider perspective on the individual and societal benefits that AI brings to the table in situations where the information reaching us is of unnecessarily low quality. We then narrow our focus to food retail where data often is abundant and of good quality—a sector ripe for transformation through AI intervention.