Office of Technology Partnerships

Do you know how much water is consumed by large language model chatbots such as ChatGPT? Researchers at UC Riverside recently found that “GPT-3 needs to ‘drink’ (i.e., consume) a 500ml bottle of water for roughly 10-50 medium-length responses” of generative chat.  In a similar vein, work done at the Vrije Universiteit Amsterdam found that by 2027, AI could consume 134 TWh of energy, roughly equal to the annual consumption of Indonesia, the world’s fourth most populous nation.  As chat-based generative AI increasingly spreads in our lives, the environmental impacts from power and water consumption, in a changing planet with resources in ever shorter supply, are becoming truly profound. 

The rapid development and deployment of AI models, tools, and agents have sparked a massive surge in investment in both hardware and software for AI products worldwide. AI training and inference calculations are primarily handled by specialized, high-performance computer chips in data centers — large facilities scattered across the country and internationally. These centers consume substantial amounts of electricity, as well as water for cooling. While there is ongoing research into resource-efficient AI solutions, the sector's rapid growth outstrips energy-efficient infrastructure.  Despite this resource consumption, AI’s promise remains great, with many positive impacts on society in areas such as healthcare, education, and agriculture.  

The University of California has long been a leader in AI development, recently gathering over 100 researchers to discuss its future. Research universities like UC Riverside continue to play a key role in addressing today’s challenges by driving technological advancement while considering societal and environmental impacts. Research universities foster innovation that supports economic growth and aligns with the public good. Through interdisciplinary research in fields like economics, technology, environmental science, and policy, universities develop solutions that enhance economic competitiveness, inform science-based policies, and prepare a skilled workforce for a shared future. Today, balancing AI’s positive impacts with its potential downsides requires a deep understanding of technology, its future directions, and the needs of local populations.  Research universities are uniquely equipped to fill this role. 

If we rush into AI without careful analysis, we risk not just environmental and human consequences but also financial ones, including widespread job loss. MIT economist and Nobel laureate Daron Acemoglu, in collaboration with Boston University, found that introducing robots into a geography led to both lower employment and wages for the human workers. In this landscape, universities play a crucial role in steering AI research away from a purely profit-driven focus. Instead of asking only what technology can do, research universities are helping champion ethical, knowledge-driven development that prioritizes skill-building and shared prosperity for all members of society. 

By developing technologies that enhance access, resilience, sustainability and fairness, while minimizing environmental harm, research universities are helping lead the shift away from outdated economic models and toward future-facing innovation. The environmental and social impact of AI demands an ethical, holistic, and sustainable approach. Initiatives like UC Riverside’s SoCal OASIS hub exemplify this, fostering collaboration among academia, government, industry, and local communities to drive research, entrepreneurship, and workforce development—creating solutions that benefit society, spur economic growth, and generate quality jobs. 

You may know that the term “Luddite” is derived from the name of the fictional Ned Ludd.  While the original Luddites were seen as being anti-technology, they can also be seen as campaigners for employee rights, protesting the job losses and poor conditions caused by early mechanization. The actions of the early industrial Luddites might be seen not as a rejection of technology, but instead a call for fair wages, job security, and humane treatment for rank-and-file workers. Luddites believed machines should not replace skilled labor or dehumanize workers, and their legacy reflects a desire for balance between progress and employee protection. This perspective serves as a reminder that technological advancement, as seen in the Industrial Revolution and now seen in the AI revolution, does not guarantee widespread prosperity but depends on its implementation and management. 

AI offers both promise and risk, driving innovation in healthcare and education while threatening jobs and reinforcing biases. Research universities are promoting responsible AI advances by educating future engineers, data scientists, and professionals in STEM and humanities on its social and environmental impact.  Universities are also helping raise awareness, support informed decisions, and expand access by collaborating with industry and overlooked communities, helping bridge the digital divide.  Toward this end, yet another UC Riverside initiative, the UCR RAISE Institute is taking a comprehensive and interdisciplinary approach to research and education in AI, from developing foundational AI technologies to advancing applications in biomedical, agriculture, and environmental sectors to drive impact and real-world solutions. Central to UCR RAISE goals is to advance AI technologies that are ethical, transparent, and beneficial to society. 

When implemented wisely, AI has the potential to bring transformative benefits to humanity. However, its advantages are not automatically shared by all. Research universities play a crucial role in driving ethical innovation, ensuring AI enhances both societal well-being and economic growth. By supporting local research institutions, we can empower them to lead collaborations between academia, industry, and the public —shaping an AI-driven future that benefits everyone, not just a privileged few. Let’s invest in the future by championing the institutions at the forefront of responsible AI development. 

Now, more than ever.

Rosibel Ochoa PhD is Associate Vice Chancellor Technology Partnerships at UC Riverside.

Shomit Ghose is Entrepreneur in Residence and Advisor at UC Riverside’s Office of Technology Partnerships, and a Lecturer at UC Berkeley’s College of Engineering.  He writes frequently on technology, including the article “Reducing AI’s Climate Impact: Everything You Always Wanted to Know but Were Afraid to Ask”.