Very true. But it doesn't stop there. One of my pet peeves is that people who believe they understand exponential growth will extrapolate too far in the future. More often than not, processes working against the exponential growth, negligible small in the beginning, will grow even faster and will reduce the grow rate, even without external intervention. In the example of COVID, it's the chance of interaction of an infected with an uninfected, even without social distancing, just because so many people will be infected, or dead. This effect is completely negligible in the beginning, but dominates in the end. (Of course, all the people modelling the epidemic have this effect in.)