Igor Khavkine (Utrecht) gives a talk at the MCMP workshop "Quantum Gravity in Perspective" (31 May-1 June, 2013) titled "Gravity. An exercise in quantization". Abstract: The quantization of General Relativity (GR) is an old and chellenging prob- lem that is in many ways still awaiting a satisfactory solution. GR is a partic- ularly complicated field theory in several respects: non-linearity, gauge invari- ance, dynamibal causal structure, renormalization, singularities, infared effects. Fortunately, much progress has been made on each of these fronts. Our under- standing of these problems has evolved greatly over the past century, together with our understandig of quantum field theory (QFT) in general. Today, the state of the art in QFT knows how to address each of these challenges, as they occur in isolation in ohter field theories. There is still an active research program aiming to combine the relevant methods and apply them to GR. But, at the very least, the problem of the quantization of GR can be formulated as a well defined mathematical question. On the other hand, quantum GR also faces a different set of obstacles: timelessness, non-renormalizability, naturality, unification, which reflect, not its technical difficulty, but rather the aesthetic and philosophical preferences of practing theoretical physicists. I will briefly discuss how the technical state of the art and a scientifically conservative philosophical position make these obstacles irrelevant. Time per- mitting, I will also briefly touch on some aspects of the state of technical state of the art that have turned the quantization of GR into a (still challenging) exercise: covariant Poisson structure, BV-BRST treatment of gauge theories, deformation quantization, Epstein-Glaser renormalization.