Finding the balance between sustainability and innovation

The possibilities generative AI creates for individuals and enterprises alike are endless. Enterprises can improve efficiency, free up valuable resources, and create a more sustainable infrastructure. However, the technology utilized to train and use the generative AI models requires significant energy and produces carbon dioxide emissions. As an enterprise leader, how do you balance improving technical offerings and sustainability in our new generative AI-driven world?

Finding the balance starts with understanding the price of generative AI on the environment, using the right tools to assess your technology, and making the necessary changes to maintain a sustainable enterprise.

The price of generative AI on the environment

A 2019 study conducted by researchers at the University of Massachusetts at Amherst estimated that 626,000 pounds (about 283948.59 kg) of carbon dioxide were produced by training a deep-learning model, which is equivalent to five cars’ lifetime emissions. While this statistic is alarming, the true risk to the environment stems from arguably one of the greatest benefits of generative AI, the generality, because of how much energy the systems require and the amount of carbon they emit.  Imagine how much carbon an enterprise needs to offset multiple AI-model trainings.

Energy-intensive data centers, from artificial intelligence and cryptocurrencies, are another new source of higher energy consumption. The International Energy Agency (IAB) predicts the electricity consumption from these data centers could double by 2026. Many of the industry’s leading technology companies that pledged to be carbon-free are now backtracking because of their evolving artificial intelligence-based business goals.

As artificial and generative artificial intelligence becomes more advanced, the toll it takes on the environment will continue to grow. However, in PwC’s 2023 Emerging Tech Survey, just 22% of business leaders cited sustainability impact as a top issue in genAI deployment. Why aren’t more people concerned? This situation presents a constant challenge for enterprise leaders who want to reap the benefits of generative AI and for IT service and solution providers who are developing these technologies. 

However, enterprise leaders and service providers can limit environmental risks. 

Limiting the environmental risks of AI models

There are tools enterprises can use to gauge sustainability and improve it. Life cycle assessments (LCAs) and Product Carbon Footprints (PCFs) provide valuable input into operations and can enable the development of sustainable products. Additionally, green AI maturity assessments and standards enable organizations to understand their current state and start to map how they can improve green maturity in AI adoption.

Enterprises can manage their environmental impact by minimizing the computational cost of genAI models through efficient algorithms, assessing the hardware requirements, and measuring the complete life cycle.

Generative AI’s positive impact on sustainability

It’s impossible to overlook the positive impact generative AI tools have on sustainability. Sustainability is complex; it involves connecting multiple ecosystems and enabling real-time data for meaningful decisions. Essentially, the same technology that is increasing energy consumption and water energy usage in data centers can also measure an organization's energy usage. 

Generative AI helps address data quality, accuracy, and consistency challenges. GenAI can also enhance supply chains and identify sustainability risks, specifically climate and social risks. It can improve logistics by reducing excess inventory and predicting demand, lowering the carbon footprint of transportation and storage. The tools make it easier for businesses to automate complex tasks across digital ecosystems at speed. Automation improves efficiency and frees up valuable resources. AI also revolutionizes ESG investing, harnessing data to understand growing consumer demands and preferences around green and ethical products. AI helps in sustainable practices and enabling interventions in the value chain. It streamlines data collection and creates awareness,  allowing enterprises to strategize more effectively through the power of data.

Generative AI enables enterprises with more accurate insights into these markers to track against net-zero goals and Science Based Targets initiative (SBTi) targets. Enterprises can use the technology for data assurance to analyze data, ensure credibility, and create a near real-time view to compare the technology’s benefits with its impact on the environment. ESG data is scattered and is available in multiple formats and languages. GenAI and frontier technology solutions provide a platform to address the critical sustainability challenges companies face today and empower holistic decision-making.

Generative AI can help enterprises optimize resources, reduce waste, and streamline operations, ultimately lowering energy consumption. It can also support sustainable sourcing by assessing suppliers and ensuring they meet sustainability criteria, providing more transparency on materials and products. The technology enables sustainable business operations through predictive maintenance, which anticipates equipment failures to reduce downtime and extend machinery life, and climate impact analysis, which allows companies to make informed, sustainable decisions by analyzing environmental, social, and governance factors. The impact on carbon emissions depends on how these technologies get integrated into broader sustainability strategies, with practices implemented during the design, development, and deployment of AI models.

Balancing sustainability with technological innovation

Enterprise leaders must find the balance between becoming more sustainable, the transformative technology they use, and the solutions they make for clients. Assessing the sustainability of distinct business areas will help gather insights to monitor energy consumption and make a measurable change. However, there is hope. According to a World Economic Forum (WEF) report, digital solutions have the potential to reduce emissions by 20% by 2050. 

With the right strategies and technologies, enterprises can find the balance between a more sustainable future and technological innovation and efficiency.