Advances in fundamental biology are providing new insights into the causes of disease. To capitalize on these findings, and develop new therapies, we need to be able to find molecules that can interact with the main actors which, most commonly, are proteins inside cells. However, for many proteins, the conventional discovery methods fail because the artificial chemical structures used are ill-suited for interacting with these proteins, and cannot bind them tightly. To overcome this limitation, we propose to copy features from natural molecules that do have the right properties. We will develop experimental methods that can synthesize trillions of these molecules — a number exceeding the count of stars in our galaxy—and rapidly identify the most promising, tight binding examples. These molecules hold immense potential, not only as tools for advancing our understanding of basic biology, but also as starting points for developing future therapeutics.