Wykorzystanie sztucznej inteligencji na rzecz równości klimatycznej: Misja Priyi Donti na rzecz uzdrowienia planety

A Lifelong Journey Driven by Early Inspiration

Priya Donti viewed the world from two distinct lenses while she was growing up. Raised in Massachusetts, her frequent family visits to India broadened her perspective, highlighting the stark differences between her life and those of many others globally. The contrast stirred a deep-seated desire within Donti to make an impact, a desire that gained momentum during her high school years. Her biology teacher’s fascinating discussions on climate change and sustainability proved transformative, firing up a sense of purpose in Donti as she grasped the prolific impact that the global environmental crisis could have on existing societal inequalities.

The Confluence of Climate, Technology and Policy

During her early academic journey, Donti aimed to study chemistry or materials science with a specific focus on advancing solar technologies. However, her time at Harvey Mudd College led her to discover an intense, newfound interest in computer science. Particularly captivating was a research piece from the UK, which underlined the significant role artificial intelligence (AI) and machine learning can play in incorporating renewable energy into power grids. The conjunction of her passions stoked an enthusiasm in Donti, steering her towards a unique academic path.

Pursuant to her passion, Donti sought a PhD at Carnegie Mellon University, designing a one-of-a-kind program that melded computer science and public policy. Her compelling research focused on building machine learning algorithms that could manage renewable-heavy power grids, displaying her determination to not only develop new technologies but also ensure they met the practical needs of the world, aligning her work closely with energy experts and stakeholders.

Even further up the road, Donti co-founded Climate Change AI, a nonprofit organization during her graduate study. The aspiration? Creating a shared platform to foster cross-disciplinary knowledge, collaboration, and education to combat global climate challenges, using AI. In Donti’s words, “In the climate space, you need experts in technical and social science tool kits, policymakers, affected users—all of them—to make a real impact.”

Pioneering Work and Aspirations at MIT

In 2023, Donti stepped into her role as the Silverman Family Career Development Professor in the Department of Electrical Engineering and Computer Science (EECS) at the esteemed MIT, enjoying a dual appointment with the MIT Schwarzman College of Computing. Donti also became a principal investigator at MIT’s Laboratory for Information and Decision Systems (LIDS). Lured by MIT’s commitment to utilizing technology for societal good, Donti felt right at home, underscoring the importance of coordinating technology, society, and policy to maximize long-term impacts.

Donti’s groundbreaking research at MIT focuses on the use of deep learning to enhance prediction, optimization, and control of electric power systems laced with renewable energy. Some of her notable work includes devising a new algorithm that enables grid operators to optimize for cost while remaining mindful of the physical constraints of the grids. Impressively, this new invention has shown to perform up to 10 times faster and more cost-effectively than previous methods, generating substantial industry interest.

Apart from her innovations in grid operations, Donti is also working on creating alternative synthetic datasets and benchmarks for training AI models in power system optimization. These synthetic data are complex enough to stimulate progress yet ensure security, addressing concerns surrounding the accessibility of real-world data due to proprietary protections.

Throughout her career, Donti’s significant contributions have garnered widespread recognition, including prestige accolades such as the U.S. Department of Energy Computational Science Graduate Fellowship and the NSF Graduate Research Fellowship. She was also featured on MIT Technology Review’s “35 Innovators Under 35” in 2021 and in Vox’s “Future Perfect 50” in 2023.

Looking to the future, Donti is eagerly preparing to co-teach a new course in Spring 2025 called “AI for Climate Action”, alongside EECS assistant professor Sara Beery, and Earth science expert Abigail Bodner. This course aims to explore how AI can be applied to improve biodiversity, ecosystems, and climate science, with Donti expressing immense enthusiasm for the collaboration involved.

At MIT, Donti has found a meeting of minds that share her commitment to impactful work against measures of success such as publications and citation counts. To Donti, the end-goal is broader – ensuring that their work makes a substantial impact on society.

What started on the gritty streets and rural landscapes of India, has now unfurled into a comprehensive career focused on leveraging AI to structure an equitable and sustainable future. For Priya Donti, the journey continues.