Our understanding of the way in which rocks record the geomagnetic field is based on an analytical theory which makes the assumption that particles are uniformly magnetized (i.e., they are single domain, SD). But, it has long been known that most rocks contain grains referred to as pseudo-single-domain (PSD); these are too big to be uniformly magnetized, and are unsuited for analytical theory. Recent breakthroughs in numerical modeling have opened new avenues for understanding the nature of PSD grains. Making it possible to estimate the temporal and thermal stability of PSD particles. This approach has gained new urgency given recent results on PSD samples which call into question their reliability for paleointensity research. The opportunity now exists to combine experimental and numerical approaches for a radical new understanding of paleomagnetic recording, with the potential to transform how paleointensity research is done.