A choice of where to put the motion
Here is a homely analogy. You are sitting on a train, looking out at a station. Are *you* moving, or is the station sliding past you? Physically the question is silly — only the relative motion between you and the platform is real, and that is the same either way. Quantum mechanics offers exactly this kind of choice. A measurement always involves two ingredients: the state (the system's condition) and the observable (the property you are asking about, like position or energy). To describe how a measurement changes over time, you can let the state carry the motion, or let the observable carry it. The split is yours to make.
These different bookkeeping choices are called pictures. They are not rival theories and they do not disagree about a single experiment. They are different desks at which to do the same arithmetic, and you pick whichever makes a given problem least painful. The thing they must all reproduce is the expectation value — the average result you would get from many measurements. Get that right and the picture you used is just a matter of taste.
The Schrödinger picture: the state moves
The default, and the one beginners meet first, is the Schrödinger picture. Here the state does all the moving — the wavefunction flows and reshapes over time, exactly as the earlier guides described — while the observables sit still, like fixed measuring instruments bolted to the bench. You hold up the same ruler and the same energy-meter at every moment, and watch the evolving state drift past them. This is the most intuitive setup: things change, the rulers do not.
The Heisenberg picture: the observables move
Now flip it. In the Heisenberg picture, the state is frozen — pinned to whatever it was at the start — and instead the observables do the evolving. The ruler itself changes with time. This sounds upside-down, but it has a wonderful payoff: the equations for how an observable evolves look almost identical to the equations of ordinary classical mechanics. A quantity like position obeys a quantum equation of motion that echoes Newton's, which makes the bridge between the quantum and classical worlds delightfully visible. Physicists working on dynamics, and especially on quantum fields, often prefer this desk.
The interaction picture: split the difference
The third desk is a clever compromise for the most common real-world situation: a system you understand well that is being gently poked by something extra — an atom sitting in a beam of light, say. Split the energy into two parts: the easy, well-understood part, and the small disturbance. The interaction picture lets the easy part ride along inside the observables (Heisenberg-style) while the state only has to respond to the small disturbance (Schrödinger-style). You park the boring, predictable motion out of the way so you can watch the interesting effect of the poke all by itself.
This picture is the natural home of the next two guides. When we ask how fast a system jumps between states, or how an atom trades energy with light, we almost always set up the problem in the interaction picture, because it isolates exactly the part we care about. So it is worth holding onto the headline: pick the picture that hides the motion you do not care about and exposes the motion you do.