This talk is divided into two related parts. First, we will call your attention to a basic, but often overlooked worrying fact, and presents ways of dealing with it. The fact is: an enormous number of galaxies in surveys like the SDSS have emission lines which are too weak (low S/N) to be classified by usual schemes (ie, diagnostic diagrams). It turns out that most of these are AGN-like, so ignoring them on the basis of low S/N (which most people do) leaves as much as 2/3 of these emission line galaxies unaccounted for. The solution: We present a number of alternative methods to rescue this numerous population from the classification limbo. We find that about 1/3 of these weak-line galaxies are massive, metal rich star-forming systems, while the remaining 2/3 are more like LINERs. In the second part, we revisit the old idea by Binette et al (1994) that post-AGB stars can account for the emission line properties of some galaxies. A "retired galaxy" model is presented and compared to data in the SDSS. We find that about 1/4 of the galaxies classified as LINERs in the SDSS are consistent with this model, where all ionizing radiation is of stellar origin. More dramatically, nearly 100% of weak-line LINERs are perfectly consistent with being just retired galaxies, with no active nucleus. If these ideas are correct, contrary to current practice, relatively few LINERs should be counted as bona fide AGN.