Do you not think it a matter worthy of lamentation
that when there is such a vast multitude of them [worlds], we have
not yet conquered one?
— Alexander the Great
The many-worlds interpretation is an interpretation of quantum mechanics, first proposed by the American physicist Hugh Everett III in 1957, according to which, whenever numerous viable possibilities exist, the world splits into many worlds, one world for each different possibility (in this context, the term "worlds" refers to what most people call "universes"). The phrase "many worlds" was first used by Bryce DeWitt, who wrote more on the topic following Everett. In each of these worlds, everything starts out identical, except for the one initial difference; but from this point on, they develop independently. No communication is possible between the separate universe, so the people living in them (and splitting along with them) would have no idea what was really going on. Thus, according to this view, the world branches endlessly. What is "the present" to us, lies in the pasts of an uncountably huge number of different futures. Everything that can happen does happen, somewhere.
Until the many worlds interpretation, the generally accepted interpretation of quantum mechanics was (and perhaps still is) the Copenhagen interpretation. The Copenhagen interpretation makes a distinction between the observer and the observed; when no one is watching, a system evolves deterministically according to a wave equation, but when someone is watching, the wavefunction of the system "collapses" to the observed state, which is why the act of observing changes the system. The Copenhagen interpretation gives the observer special status, not accorded to any other object in quantum theory, and cannot explain the observer itself, while the many worlds hypothesis models the entire observer-observee system.