How to balance the ecological dangers and promises of generative AI

The rapid adoption of ChatGPT has elevated the negative environmental impacts of generative AI from a hyped new technology to an immediate concern for organizations. Many seemingly good use cases powered by this emerging technology will do more harm than good in terms of greenhouse gas (GHG) emissions and electricity and water consumption.

Nevertheless, generative AI can also accelerate positive sustainability and financial outcomes. If used properly and under human supervision, the technology has the potential to help companies reduce sustainability risk, optimize costs and drive growth.

To balance the dangers and promises of this technology, organizations must do two things. First, we need to recognize and reduce the energy footprint of generative AI to make it more environmentally friendly, and then identify, evaluate and prioritize the use cases for environmental sustainability.

Recognize that generative AI has a consumption problem

Generative AI relies on massive models trained with massive amounts of data, leaving it thirsty for cooling water and hungry for electricity. In some cases, the technology can consume enormous amounts of both. Although electricity-related greenhouse gas emissions will decrease in the long term as more renewable energy sources are used, more powerful generative AI models will require greater computing capabilities.

The problem of technology-related electricity and water consumption goes far beyond generative AI. Gartner predicts that 75% of executives will face technology-related electricity constraints by 2025. As the needs of technology and society increasingly compete, CIOs do not want to find themselves in the position of competing with local communities for finite resources. .

Reduce the energy footprint of generative AI

To make generative AI more environmentally friendly, it must be as efficient as the human brain. One of the reasons why the brain is so energy efficient is that it organizes knowledge into network structures. The current equivalent of this is composite AI, which uses similar network structures and techniques to complement the current brute force, deep learning method.

Generative AI must also be put on an electricity and water diet. Stop training AI once the improvements level off. Keep model training data local, reuse models that have already been trained, and use more energy-efficient hardware and networking equipment. Balance data center ‘follow the sun’ workloads that are better for clean energy production with ‘unfollow the sun’ measures that are better for water efficiency.

Another way to make generative AI more environmentally friendly is to do it in the right place and at the right time. The carbon intensity of local energy supplies varies due to a number of factors. The best practice is to use energy-aware job scheduling, along with carbon emissions tracking and forecasting services to reduce associated emissions.

Also try to buy new clean energy where you want to use it. The Greenhouse Gas Protocol considers requiring companies to provide a more detailed analysis of clean energy by location, time, or both.

Identify potential use cases for environmental sustainability

There are three broad areas where generative AI use cases can accelerate environmental sustainability: mitigating risk, optimizing costs, or driving growth.

Regulatory compliance is one way generative AI can help your organization mitigate environmental risks by identifying and interpreting relevant sustainability laws, standards, guidelines and reporting requirements, including updates over time. It can develop an action plan to achieve compliance and training materials to educate employees on specific regulations.

From a cost optimization perspective, generative AI can help support decision making. It can analyze internal sustainability data and identify patterns, trends, areas for improvement, feasibility, risks and benchmarks. It can provide insight into how organizational decisions will affect sustainability and predict likely future performance. Companies can therefore plan and select optimal pathways to achieve greenhouse gas emissions reduction targets.

Generative AI can also be used to drive sustainable growth by applying it to discover alternative resources and materials. It can provide suggestions for sustainable alternatives to conventional inputs; insights into technological innovation such as nanomaterials; and information on availability, performance and environmental impact.

When considering generative AI use cases for sustainability, it is important to evaluate the positive and negative impacts. Look at the positive business value in terms of financial and sustainability benefits, as well as feasibility, and negative environmental impacts as measured by greenhouse gas emissions and electricity and water consumption.

You can then prioritize your investments based on three levels: invest now; first reduce risks and energy consumption; or not invest. That way, you’ll use generative AI to accelerate your organization’s positive sustainability outcomes by using only those use cases that create more value than they destroy.