Abstract: Dark matter has been one of the central puzzles in physics for nearly a century, and after four decades of direct-detection experiments, with sensitivities improving by roughly ten orders of magnitude, we still have no definitive detection. As the search continues, along with all the efforts in improving detector technology, the progress also requires pushing in two complementary directions: understanding more precisely what signal we should expect and expanding where we can look. I will first introduce our work using the FIRE cosmological simulations to explore how the Milky Way’s merger history shapes the local dark-matter velocity distribution, an important astrophysical uncertainty for direct detection, and shortly comment in light of the recent intriguing LZ event. I will then turn to the low-mass frontier, where conventional detectors rapidly lose sensitivity, and present a new approach using the one-dimensional plasmon mode in carbon nanotubes to search for light dark matter. A better understanding of the Milky Way’s recent merger history, especially the impact of the LMC, will be increasingly important for interpreting future searches, while rapid advances in condensed-matter systems and sensor technologies are opening an exciting new era for light-dark-matter detection.