How does the science behind baking a gingerbread house relate to the progressive world of artificial intelligence? If you ask Miranda Schwacke, a PhD student in the Department of Materials Science and Engineering at MIT, she’ll tell you that the link lies in the concept of how structure influences function. It’s an intersection of delicious biscuits and intelligent brains – believe it or not.
Ein Mitglied von Küchenangelegenheiten, a group comprising MIT grad students that turn cooking into a platform for explaining scientific concepts, Miranda didn’t just set out to bake a confectionary delight. She was after a gingerbread house that was not just appetizing, but also structurally durable. It culminated into a hunt for a foolproof recipe by experimenting with differing cookie formulations, which eventually boiled down to manipulating a key component: butter.
“Butter houses water and turns it into steam when baked, creating air pockets. I presumed that reducing the butter would make the cookies denser and stronger, perfect for building,” she explains. This very experiment found its place in a Kitchen Matters video demonstrating how the internal structure of a material, like cookie dough, can drastically influence its characteristics.
Miranda has always been captivated by how materials behave under different circumstances. This interest laid the foundation for her groundbreaking research on neuromorphic computing – an energy-efficient approach to AI inspired by the intricacies of the human brain. Unlike traditional computing that separates memory and processing, neuromorphic devices fuse these two functions just like the human synapses do. Miranda says, “Training large AI models consumes massive amounts of energy. Our brains, on the other hand, can learn consuming far lesser energy. The efficiency of our brains forms the core idea behind neuromorphic systems.”
Starting with her earliest memories of growing up with a marine biologist mother and an electrical engineer father, Miranda’s fascination for science was fostered as a way to decipher the world. In high school, she uncovered her love for materials science and completed her bachelor’s degree at Caltech where she pursued studies on nanostructured materials and electrochemical systems such as battery and fuel cells. Today, at MIT, she’s putting in the hard work to innovate for energy-efficient AI, focusing precisely on how magnesium ions interact with tungsten oxide – a material whose electrical resistance can be meticulously regulated, to mimic synaptic behavior.
Darüber hinaus ist es Miranda ein Anliegen, einen nachhaltigen Beitrag zur breiteren wissenschaftlichen Gemeinschaft zu leisten. Sie engagiert sich ehrenamtlich, um Kinder mit praktischen Experimenten zu unterrichten, und hat in verschiedenen Positionen in wissenschaftlichen Organisationen gearbeitet. Miranda ist davon überzeugt, dass eine effektive Kommunikation für ein kollaboratives Forschungsumfeld entscheidend ist. Geleitet von ihrem Streben nach wissenschaftlicher Neugier hofft sie, Professorin zu werden und die nächste Generation von Wissenschaftlern als Mentorin zu unterstützen, um sie zu inspirieren, Herausforderungen anzunehmen und bahnbrechende Entdeckungen zu machen.
Diese Website verwendet Cookies.