Light harvesting and light-sensing biosystems are fraught with noise. Despite this, biosystems operate at the extremes of temporal response, single-quanta sensitivity, and light conversion efficiency. Perhaps unsurprisingly, a complete physical understanding of Nature's striking performance is lacking. In this talk, I present new physics of biological vision and light-harvesting systems, with a particular focus on the fundamental aspects of statistical noise. As a first application, I explore the infrared vision in pit vipers. I show that a vision system optimized to assess thermal fluctuations at each pixel in space is remarkably efficient at finding boundaries between regions of differing temperatures. Beyond explaining several long-standing questions in the pit viper community, this work has the potential to expose new understanding in biological vision systems, many of which show evidence for single photon sensitivity. I then present a quantum physicist's perspective on photosynthesis, as first laid out by several pioneers including Erwin Schrödinger and George Gamow. By understanding the connection between light harvesting network structure, noise, and power conversion, I show how photosynthetic antennas can be finely tuned to maximize light harvesting efficiency. If, as on Earth, photosynthetic life on exoplanets evolved to optimize its light harvesting quantum efficiency, then the colors of reflected light from such a surface biosphere may be detectable as a surface biosignature. Based on this key premise, I give an update on our active work in identifying photosynthesis on the surface of distant exoplanets. The work presented in this talk - which identifies general principles of noise in Nature - promises to have applications across various disciplines ranging from quantum nanoscience and computing to bionanoscience and astrobiology.