Achuta Kadambi, an assistant professor at the UCLA Samueli School of Engineering, published a column in the journal Science about how medical devices can be fundamentally biased—not just in dataset representation as has been widely reported, but from a deeper root: the laws of physics.