That's the idea that probably contributed to this problem: the model says that ring 0 is all powerful. But in a richer model, we might change who has ring 0—we

That's the idea that probably contributed to this problem: the model says that ring 0 is all powerful. But in a richer model, we might change who has ring 0—we might sell a machine, or get it infected and then want it clean. Root kits that can persist even when a later ring 0 wants them gone? That's surprising. So we learn to enrich our model to say something about time and causality, maybe.