This is the AI balancing act: between its huge potential and growing emissions

The global artificial intelligence (AI) market is projected to post a CAGR (compound annual growth rate) of 38.1% and reach a value close to $1,600 billion by 2030 — a meteoric rise aided as much by big data as it is by software and hardware.

According to Statista, the amount of data today’s digitised economy creates is growing annually by 40% and is expected to reach 163 trillion gigabytes by 2025, which will further fuel growth in AI.

The data value chain involves devices and sensors to capture the data, networks for communicating data and data centres to store them. All of this requires natural resources and energy to build and transport the devices and products, which emit greenhouse gases throughout their lifecycle. Beyond the products themselves, energy is also required to run machine-learning algorithms.

Data centres are responsible for 20% of all electricity consumption in Germany’s financial hub Frankfurt. Neural networks — computing systems inspired by the human brain and used by most modern AI algorithms — are particularly power-hungry.

The general trend in AI is toward ever more power consumption because of the computing used to train the model and the computing used to infer new data from those models. According to OpenAI, creators of ChatGPT, the computing used to train the average model increases by a factor of 10 each year. Some believe machine learning is on track to consume all the energy that can be supplied.

Many studies have investigated how data and AI can be used to solve sustainability issues by revealing new insights and contributing to smarter decisions. Yet some of those benefits are offset by rebound effects and human practices, and the direct negative impact of the data and AI value chain is significant.

“AI will transform business practices and industries and has the potential to address major societal problems, including sustainability. Degradation of the natural environment and the climate crisis are exceedingly complex phenomena requiring the most advanced and innovative solutions,” argue AI researchers Rohit Nishant, Mike Kennedy and Jacqueline Corbett.

At a corporate level, they propose five ways to reconcile AI’s impact on the environment and its centrality to a solution to the climate crisis.

1. Adopt a multilevel view to capture the complexity of the real world, and limit rebound effects, for example.

2. Use a system dynamics perspective to capture interactions and feedback loops among the technology, users, and other stakeholders.

3. Follow a design thinking approach to minimise potential unintended consequences and improve the effectiveness of AI solutions.

4. Understand the psychological and sociological underpinnings of human response for effective long-term solutions.

5. Examine the economic value of AI for sustainability to develop our understanding of how AI differs from conventional IT.

“The true value of AI will not be in how it enables society to reduce its energy, water and land use intensities but rather, at a higher level, how it facilitates and fosters environmental governance,” the researchers add.

Various reports have urged governments to pursue sustainable AI and one of them suggested the following.

As the volume of data and the use of AI increase and spread across industries, company sizes and countries, corporations will be challenged increasingly on the environmental implications of data and AI by the public and most likely, regulators.

With an increased maturity level within the public, companies and governments, only counting the benefits of using AI for solving environmental issues will no longer suffice. Balancing with other negative impacts of the organization (on topics other than digital) and taking into consideration the direct negative impact will be necessary.

Achieving that measurement will require a systemic view of environmental impacts, way beyond its purely digital impact (whether direct or indirect), and including the stakeholders outside the company (clients and suppliers for example).

Aiming to have a positive effect will require leveraging expertise that was either absent or isolated from the engineering expertise (behavioural science and design thinking, for example).

From a competition dynamics perspective, it is likely that a few companies will act above the standards and expectations and build a distinctive positioning. For those that do, however, the upside will be about branding and a way to support a premium positioning. However, it will also require hard commitments on many dimensions to avoid the greenwashing label, and so will entail significant changes to their business model.

Even fewer companies will put the environment first and redefine their operations and models accordingly— an approach that would require them to seriously consider whether they need all the data they have access to and if they need energy-intensive models.

Some companies, more numerous, will define strategies for radical energy savings for cost reasons. The higher the relative importance of energy in the costs, the more efficiency constitutes an advantage. Similarly, some companies will try to secure cheaper access to clean energy, as big tech firms are already doing.

How we balance AI’s innovation potential with its environmental impact is a difficult yet necessary conversation to have. It is a necessary conversation because every day we are reminded of the urgency of the decisions to make. It is a difficult one because it is not a purely technological issue — it is also one of governance.