The Many-Worlds Interpretation has a reputation for being exotic and science-fictional, which is understandable. It does, after all, claim that every quantum event spawns an entire new universe. But physicists like David Deutsch and Sean Carroll argue that it is not exotic at all, it is the conservative interpretation, the one that takes the mathematics of quantum mechanics at face value without adding anything.
Here is the core claim stated carefully: there is one quantum state of the universe, the universal wave function. It evolves deterministically and continuously according to the Schrödinger equation. No collapse, no special role for observers, no mysterious exception for measurements. When a quantum event with two possible outcomes occurs, the wave function of the universe evolves into a superposition of both outcomes. Because the measuring device and the observer become entangled with the quantum system, the observer also enters a superposition: one version sees outcome A, another sees outcome B. These two versions of the observer quickly decohere, become informationally isolated from each other, unable to interfere with each other, and from that point forward they are, for all practical purposes, separate worlds.
The analogy that makes decoherence vivid: imagine a river forking. Once the two streams have separated far enough, water from one cannot easily flow into the other. The branches of the wave function are like river branches that have diverged too far to reconnect. Each branch contains a complete, internally consistent world.
Philosopher David Wallace, in The Emergent Multiverse (2012), has developed the most rigorous contemporary version of Many-Worlds. His key claim: the branching structure of the multiverse is emergent, it is not written into the fundamental physics as a discrete set of worlds, but arises from the mathematics of decoherence just as tables and chairs emerge from atomic physics. The worlds are real, but they are real in the way that biological species are real, emergent patterns in a more fundamental substrate.
The philosophical case for Many-Worlds rests on what Deutsch calls the Occam's Razor argument. Every other interpretation of quantum mechanics adds something to the bare formalism, collapse postulates, hidden variables, special roles for observers. Many-Worlds adds nothing. It just takes the Schrödinger equation seriously, all the time. If you believe quantum mechanics is true and don't want to add extra structure, you get Many-Worlds automatically. The price you pay is a lot of parallel branches. The question is whether that ontological extravagance is a cost or a consequence.
Sean Carroll puts it bluntly: "The Many-Worlds Interpretation is not a hypothesis, it is what you get when you take quantum mechanics seriously." The other interpretations are attempts to avoid the conclusion. Whether that conclusion should be avoided is the philosophical question.