First, a brief description of Bell\'s theorem will be given.
It states states that there is no classical-like (local realistic) description of all quantum predictions. Next, a plausible class of non-local realistic models will be presented which is incompatible with quantum mechanics, as first shown by Leggett. Experiments confirming the incompatibility will be described. Finally, it will be argued that quantum mechanics can be seen as a theory of systems with limited information resources.