In summary, at a resolution of typically a few tens to a few hundred parsecs, the local disk galaxy population appears to host bulges that more and more resemble disks, and disks that more and more resemble bulges. Disks may be denser than exponential, and bulges may be less steep than de Vaucoluleur's structures. Bulges are claimed to have the kinematics of thickened disks, and buckled bars can be as dynamically hot as the structures claimed to be "true" bulges. The average stellar population properties (i.e., stellar ages and metallicities) remain the only surviving distinction between "massive" and "small" bulges and, more generally, between massive bulges and the centers of disks. This could certainly be an indication of different formation processes, but could also be the result of similar processes occurring at different epochs in the Universe (and thus naturally generating a positive correlation between stellar densities and ages). In my view, what is needed at this point is a shift of the debate from the arena of morphological classifications, where bulges and disks are distinct entities and the question "what is the origin of bulges" is kept distinct from the question "what is the origin of disks," to one where disk galaxies are studied as a whole without the constraints of a rigid classification scheme. The historical focus on morphology, while highlighting many details in the trees, may in fact have hidden the true nature of the forest. The expected outcome will be a renewed concept of the "Hubble sequence" that will be ultimately be based on physical rather than morphological considerations. Clarifying what we really see nearby as the endpoint of the galaxy evolution process is essential in order to meaningfully answer how the distant progenitors, which are seen in the most remote regions of the Universe, transform themselves to become the descendants that populate our own surroundings.
Acknowledgements. I thank L. Ho for the invitation to this very stimulating meeting, and especially for his patience waiting for this manuscript. I am grateful to my collaborators, V. Debattista, L. Mayer and B. Moore for kindly making available some of our results prior to publication. Many thanks to F. van den Bosch and S. Lilly for comments on a previous version of this manuscript.