SABRE, a method pioneered about a decade ago, uses parahydrogen and reversible exchange in solution to hyperpolarize organic molecules. In recent years this method has been adapted to polarize heteroatoms such as 15N, rapidly (in seconds) on several hundred different molecules, with an apparatus that can be built for about 1% of the cost a DNP system. Two different strategies have been successful: direct transfer of order from parahydrogen at about 0.5 microtesla (where the resonance frequency difference between nitrogen and hydrogen is about the same as a J coupling) and very weak irradiation of the nitrogen resonance at high field (irradiation strength comparable to a J coupling). Both of these are unusual regimes for magnetic resonance, and provide unique opportunities for exploring spin physics in underexplored domains. This is important because, while progress in a short time has been phenomenal, the fundamental limitation of this technique today is scalability; these methods typically work well at low (mM) concentrations, and not that well in water. New pulse sequences we have developed, coupled with a new quantum Monte Carlo simulation approach, have drastically improved our understanding of the spin dynamics in these complex systems. They have also led to significant signal enhancements, new imaging agents (such as para-nitrogen gas and a variety of injectable compounds) and new strategies for clinical-scale hyperpolarization.