The idea is that the cat is both dead and alive until you look, because the state of alive or dead is entangled to a quantum superposition that is both states at the same time.
That's a possible idea, but it's not necessarily true, and it is not the idea that Schrödinger wanted to communicate.
His aim was to show that extrapolating the quantum-theoretic description of state (allowing entanglement) to macroscopic systems makes that description not viable as description of the real state of the system, which macroscopic systems are assumed to have at all times; it is dubious and even ridiculous.
This is because the description allows macroscopic systems in contact with quantum systems (such as decaying nucleus) to be in strange states which are (by the very theory that predicts them) impossible to observe (because these quantum states can't be observed, observation destroys them in favor of the "sane" states), and macroscopic systems are observed all the time and such "insane" states indeed were never observed. So the very idea of applying the formalism which produces such states easily and all the time is suspect.
It is one thing to allow the assumption the electron has no definite position, and entirely another to admit that macroscopic machine has no definite state when nobody's looking. It does not matter if it is cat or Geiger counter. The trouble is with the positivist attitude that things don't exist when we're not looking. Schrödinger did not like that, and many people still don't. It is suspect to believe that things exist that can't be measured (superpositions of macroscopic systems), and that things that can be measured (he usual macroscopic states) don't exist most of the time when we're not measuring.
There is a pragmatic way out that preserves usefulness of the quantum theory - use quantum theory as generalized statistical physics, the so called "statistical interpretation". It predicts probabilities of results of measurements, with good track record. It does not necessarily say much about "what is really going on" or "what is the real state of the system" when no measurement is being made. Quantum state is a concept we use, rather than 1-to-1 image of the real state of the system. It's like probability distribution in statistical physics; a tool, a map, not the studied thing, not the territory.