Is AI a benefit or cost to those trying to reduce the environmental impact of their business?

Artificial Intelligence (AI) is becoming an increasingly common tool in business. However, its environmental impact is complex, leaving many organisations that are working to reduce their carbon emissions wondering whether AI is a benefit or a cost to the planet. While AI can support efficiency, innovation and sustainability initiatives, its energy consumption for training and running models, along with the associated carbon footprint and water usage, poses a significant environmental challenge.

In this article, we will outline the negative and positive impacts of AI so you can formulate your own perspective and ultimately assess whether AI is right for your business, and whether there are actions your organisation could take to reduce its environmental impact whilst still staying productive and competitive.

Negative Impacts:

High Energy Consumption: AI, especially deep learning, requires massive amounts of electricity for training and running models, contributing to greenhouse gas emissions, especially if powered by non-renewable sources. This means organisations should consider the energy demands of AI alongside its potential efficiency benefits.

Data Centre Infrastructure: The data centres that house AI servers consume substantial energy and water for cooling, further straining resources.

E-waste: The rapid development and deployment of AI hardware lead to increased e-waste, containing hazardous chemicals that can contaminate soil and water.

Resource Extraction: The mining and production of materials for AI hardware, such as metals and rare earth elements, can lead to environmental degradation, including soil erosion and pollution.

Water Footprint: AI data centres require significant water for cooling, putting pressure on water resources, especially in regions already facing water scarcity. It is estimated that global AI infrastructure already consumes 6 times more water than the country of Denmark (population: 6 million) and that is, of course, set to increase.

Carbon Footprint: The manufacturing and transportation of AI hardware, as well as the energy used to power AI systems, contribute to a substantial carbon footprint.

Consumption Change: As AI’s algorithms become more accurate, the targeting and timing of adverts and articles which promote greater, often unnecessary, consumption at the push of a button is an ever increasing threat to the sustainability movement’s push for people to do more with less.

Unintended Consequences: There are concerns that the development of and reliance upon AI could have unintended impacts: for example, as the number of AI powered self-driving vehicles increases, the numbers of those who currently take journeys on foot or by bike could decrease.  Additionally, on a topic which is regularly subject to ‘fake news’, AI could support the generation and proliferation of misinformation, downplaying threats or mis-reporting impacts.

Positive Impacts and Potential Solutions:

Environmental Monitoring and Management: AI can be used to monitor deforestation, track pollution levels, and optimise resource management, aiding in environmental conservation efforts.

Sustainable Agriculture: AI can help optimise irrigation, fertiliser use, and pest control in agriculture, leading to more sustainable farming practices.

Renewable Energy Optimisation:  AI can improve the efficiency and reliability of renewable energy systems, such as solar and wind power, contributing to a cleaner energy future.

Waste Management:  AI can help optimise waste collection and recycling processes, reducing landfill waste and promoting resource recovery.

Climate Modelling and Prediction: AI can be used to develop more accurate climate models, helping to understand and predict the impacts of climate change.

Sustainable Supply Chains: AI can help optimise logistics and supply chains, reducing transportation emissions and promoting sustainable sourcing.

Energy Efficiency: AI can be used to optimise energy consumption in buildings and industries, reducing energy waste and emissions.

AI can assist the development of more efficient and sustainable technologies and processes. And, with innovative design and construction, some of the environmental impacts of the growth of AI can be mitigated. Taking data centre developer and hosting provider, Deep Green, as an example, their high density computing infrastructure is located next to facilities requiring heat, the best example of which is the capture of the heat energy generated by a data centre to heat the neighbouring municipal swimming pool.

In advising organisations who are working hard to reduce their environmental impact, the Investors in the Environment (iiE) team suggest the following steps are taken:

• Develop a policy for use, considering how AI might be used in your organisation with suggestions about what is appropriate use and what is inappropriate use.

• Communicate that policy with staff, giving them a chance to input or feedback on its content and roll out.

• Monitor its use and its impact: this might involve staff or stakeholder surveys, analysis of production or service provision, comparison of resource use and the evaluation of other efficiency (ie staff hours) gains.

• Record results: make comparisons from one period to the next.

• Review and update: having analysed hard and soft metrics, consider if changes should be made and if they are, ensure staff are correctly informed and trained on those changes.

 

Weighing up the impact of AI on your organisation’s carbon footprint? Speak to a member of our team and we’ll help you understand more about the environmental costs and benefits of AI, and identify practical ways to reduce its impact. Email us at [email protected].com or book a consultation today.