Age specific mortality curves exhibit an age t c (about 10 years) which plays a crucial role in that the mortality curve decreases hyperbolically in the age interval A before t c and increases exponentially in the interval B following t c. For those familiar with reliability theory, region A is called the " burn in " phase and B is the " wear out " phase. Using the exponents of the hyperbolic and exponential phases, we introduce a new 2 dimensional map of diseases. This permits the classification of diseases into three broad classes: AS1, AS2 and S. Class AS1 includes all diseases arising from congenital malformations which dominate infant and child mortality; class AS2 includes degenerative diseases such as dementia and Altzheimer's which dominate old age mortality. In class S, which includes most infectious and metabolic diseases, the exponents from both aging phases contribute to positions on the map. Cancer is one of these mixed cases but is closer to class AS2 than AS1. A second line classification is needed to resolve S cases and to this end we introduce a 3rd dimension, namely (calendar) time. Using historical data we show that in their response to treatment (particularly vaccination), S diseases fall into three sub-classes. (i) Class E diseases (e.g. measles or meningococcal disease) which have been almost eliminated at all ages (ii) class C diseases (e.g. tuberculosis) which can be cured but whose cure becomes less effective at old age. (iii) Class U diseases for which radical cures are still unknown. Regarding the future, the fact that the wear-out process of numerous diseases already starts around the age of 25 means that a major extension of the human lifespan beyond 120 certainly also requires to uncover the secret of the " elixir of eternal youth " which has driven timeless human efforts and still seems unlikely in the foreseeable future.