Yeah. If the signal changed from 0V to +3V, in any way, at any time, then technically it is AC. However, we might treat it as DC for engineering purposes depending on the context. If the system the DC current is going into is stable and responds quickly enough, we can ignore the effect that “starting” the signal will have. However, most non-trivial circuits have non-trivial dynamics, so we need to make sure that, at a bare minimum, they can be started properly.

Dynamic circuits have, amongst other things, resistances, inductances, and capacitances. Resistors are boring; put in a current I, get back voltage V = IR. Circuits with only resistances are not dynamic. Capacitances and inductances are themselves dynamic elements. For them to exhibit any interesting behavior, the signal (input current or voltage) needs to be changing. Circuits can (but will not necessarily) experience weird behavior when turning on if the power is taken from zero to some voltage too quickly. For example, if a current in an inductor is set from zero to a nonzero constant, you will get a huge (theoretically infinite) spike in voltage. This is because voltage across an inductor is proportional to the rate at which the current signal changes. In the case of a current switched from zero to non-zero, the change from zero to not-zero happens almost instantly, implying a voltage that approaches infinity as the transition becomes steeper. Practically, you get a voltage spike which, when multiplied by current, could generate enough power to blow up whatever you’re working on.

I didn’t mention this in any of the comments, but real “DC” sources are dirty. Noise is a class of signals that can take on random values at all times. The fact that it can be a different random value at any time implies that noise is a subclass of AC signals. A “pure” source contaminated with noise is colloquially called a “dirty” source.

Dirty power supplies, the effect of turning the power on and off, and other supply imperfections are why if you look at the power sections of electrical circuit diagrams, you will see capacitors from the supply voltage to ground. The current through a capacitor is directly proportional to the rate of change of the voltage across it. A constant voltage would imply a zero rate of change, so no current is going through the capacitor. Colloquially, it shouldn’t do anything unless there is a disturbance in the circuit.

Engineers generally expect power supplies to “kinda suck”, even those marked as DC. If there are any other dynamic circuits being powered by the supply (and there almost always are; we want our electronics to actually do things for us!), they will be affected and possibly compromised entirely by a dirty power supply.

I couldn’t really think of an application where blowing could be treated as a DC signal and talking could be treated as an AC signal. From an electrical engineering position, blowing is a particular subclass of speech signals, for which there is a dearth of literature, all of which hinges on the assumption that speech signals are “AC”, e.g. time-varying.