To date, the CMB has been our cleanest probe of the physics of the early universe. It is in principle limited in its information content. In the first part of the talk, we survey the prospects for gains in our theoretical understanding of the early universe from exhausting observations of temperature and polarization anisotropies (and possibly measuring spectral distortions) in the near to mid term. In the second part, we present an amusing observation that the CMB can be used to bound (possibly even count) the hidden field content of the universe, since a large number of hidden fields can resume to potentially observable logarithmic runnings in CMB observables in the context of single field inflation, courtesy of a `large N' expansion. Thus, one can convert ever more precise bounds on absence of such effects into bounds on the hidden field content of the universe, with (marginally) interesting implications for a various phenomenological constructions.