A Greek scientist at MIT is tackling one of the biggest hidden costs of the AI boom: the massive energy waste inside dat...

A Greek scientist at MIT is tackling one of the biggest hidden costs of the AI boom: the massive energy waste inside data centers that power it all. Christina Delimitrou, an associate professor in MIT's Department of Electrical Engineering and Computer Science, recently earned tenure for research that uses machine learning to make large-scale computing systems run far more efficiently. The core problem her team identified is striking. Many large computing systems operate at just 15% of their actual capacity, even while consuming enormous amounts of electricity. Delimitrou started uncovering this during her graduate studies at Stanford, working alongside professor Christos Kozyrakis, and has been building solutions ever since. Her lab's approach uses AI itself to close that gap, developing tools that automate how computing resources are allocated in cloud environments and redesigning server architectures to support more applications with fewer physical resources. One of her team's concrete outputs is a tool called Seer, which uses deep learning to predict and prevent failures in online services before they happen, reducing outages in everyday apps like video calls and music streaming. Delimitrou grew up in northern Greece, studied computer engineering at the National Technical University of Athens, and moved through Stanford and Cornell before joining MIT's faculty in 2022. She is also a member of MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL). She told the Athens-Macedonian News Agency that AI's energy demands are one of the most important technological challenges of the coming years, and that the real innovation is not just building more powerful systems but making them efficient, accessible, and sustainable. She also credited her training at the National Technical University of Athens, saying the emphasis on critical thinking and the people she worked with there gave her the foundation to pursue research abroad. #Greece #MIT #ArtificialIntelligence