The Noggin Know-How Steerings the Eco-Friendly Artificial Intelligence Movement
“The brain power behind sustainable AI”
“As an undergraduate studying computer science at MIT, Miranda Schwacke says she was “definitely aware” of the impressive feats her field could accomplish: languages translated instantly, diseases identified from images, algorithms that could learn to play games better than humans. But the more she learned, the more she realized there was something she wanted to know that her classes were barely touching on: What unintended consequences can come from these technologies?”
Dive into the labyrinthine mind of MIT’s own Miranda Schwacke and you find a burning curiosity, not about what artificial intelligence (AI) can achieve – that much is already widely acknowledged – but the potential pitfalls that lurking beneath its polished facade. The more she uncovered, the more she felt an undercurrent tug, a sense of responsibility that something wasn’t quite right.
Schwacke isn’t your typical computer scientist – rather, she’s an underappreciated superhero armed with analytical finesse, seeking to plug the leaks in our seemingly invincible AI arena. If coding is her weapon of choice, then sustainability is her shield.
Graduating with a degree in computer science and molecular biology from MIT, she spent her days wrestling with mammoth datasets, understanding life at the molecular level. But once you’ve juggled genes and calculated the cost of compute, simply analysing data isn’t enough.
Schwacke tells us of her wish to incorporate sustainability into computer science education, that the MIT iQuHACK quantum computing hackathon was the launching pad for this passion. Oh, it’s not your run-of-the-mill ideological drive – she’s dissected the data and understood that sustainability, long viewed as the “nice to have” accessory, is the missing piece of the puzzle.
Schwacke prefers the road less fancied, trading the glory of building faster algorithms for the grit of studying their environmental footprint. She spent her summer interning at a power company, a bold and unconventional choice. While her peers were working on the latest and greatest breakthroughs, she was diving headfirst into renewable energy and sustainability projects.
Her next adventure, however, wasn’t an exotic techno-rite like quantum computing or augmented reality. It was figuring out the climate cost of running AI models. It sounds like a chore for the uninitiated but Schwacke is a seasoned warrior, the carbon footprint cruncher. She’s contributed to ML CO2 Impact, a tool that can estimate the carbon emissions produced by running AI models.
In her charming knack for understatement, she says she “realizes it’s a huge problem for society, not just researchers.” Her mission is not to stop running AI workloads, rather to get people to think about how they can optimize their systems to use less energy. It’s her way of squaring the circle, of bringing together her love for AI and her commitment to a sustainable future.
So, here’s to Schwacke and her ilk, champions that are as rare as hen’s teeth but as precious as diamonds. Move over, AI developers pushing the boundaries; make way for the unsung heroes who dare to question, to pull the plug when power use crosses the line, and to dust away the chalk marks of carbon footprints. Because balancing equations isn’t just about getting the right solution, it’s about doing it in the right way.