In many diverse physical systems the constituting blocks are well
localized quantum modes. The examples are: spin glasses, structural glasses,
strongly disordered semiconductors and superconductors, etc. In the absence
of the interaction, the locally measured spectrum of these systems would be
a set of sharp levels. What happens to the spectrum when the interaction is
switched on? Will it remain a set of localized levels or will the individual
levels broaden? In different terms, when localized modes do not get
entangled with the far away world? This is the question that condensed
matter community knew well but carefully avoided for the last 50 years.	
	There are two completely different reasons why this problem is very
important for the quantum computation: (1) implementation of QC implies the
construction of the physical system which is disentangled from the rest of
the World, (2)  level broadening of the local modes surrounding the physical
qubit lead to the noise acting on it. I will review the recent experimental
and theoretical progress in the studies of the low temperature noise in
superconducting system.