Two major developments have influenced the environment of actuarial math-ematics. One is the arrival of powerful and affordable computers; the onceimportant problem of numerical calculation has become almost trivial in many instances. The other is the fact that today's generation is quite familiar with probability theory in an intuitive sense; the basic concepts of probability theory are taught at man), high schools. These two factors should be taken into account in the teaching and learning of actuarial mathematics. A first consequence is, for example, that a recursive algorithm (for a solution) is as useful as a solution expressed in terms of commutation functions. In many cases the calculations are easy; thus the question "why" a calculation is done is much more important than the question "how" it is done. The second consequence is that the somewhat embarrassing deterministic model can be abandoned; nowadays nothing speaks against the use of the stochastic model, which better reflects the mechanisms of insurance. Thus the discussion does not have to be limited to expected values; it can be extended to the deviations from the expected values, thereby quantifying the risk in the proper sense.