So let me discuss just one alternative: suppose there is a dense cluster of stars in the centre of a galaxy hosting a QSO, and what we are seeing is the combined effect of several tens of SNe explosions. A single SN at it's peak is about as bright as a whole galaxy, so we need about 100 SNe and we're done. This model has withstood for several decades, but it does have some pitfalls. For example, we don't know how such SNe would generate these tremendous radio-lobes that some QSOs have. But then, we don't know how a SMBH does that either. As Luis Ho puts it: `our confidence that SMBHs must power AGN largely rests on the implausibility of alternative explanations'. Most of the arguments in the next sections just suggest the presence of a very massive, dense object (MDO) in the centre of galaxies (not just in galaxies harbouring a QSO - in fact, some of the best evidence is in galaxies which do not have an active QSO), but not really require the MDO to be a black hole, i.e. an object with an event horizon.