During the last decade dozens of papers have been published, whose approval for printing signifies a consent for changes in some biology handbooks. Stem cells have found their way to newspaper headlines and even became an element of presidential campaign in the United States. Indeed, an important issue on scientific and medical ground is that recently several paradigms in hitherto force in developmental biology have been challenged. First, the paradigm of barrier between embryonal layers has been questioned. Second, it has been demonstrated that even in such structures as the central nervous system and heart muscle there is a steady but distinct renovation, even in adult mammals. Third, new models of cell differentiation have been proposed. For many years, developmental biology was based on instructional models of differentiation. At present, the instructional model is increasingly frequently replaced by the stochastic model. Furthermore, the question of debate is whether there is still any rationale for further propagation of hierarchic models. Current dispute concerning the above mentioned concepts and paradigms, which is currently taking place among biologists studying the development of organisms, has a direct transmission onto practical applications thereof. For example, lack of barrier between embryonal layers (or its permeability) paves the way for the use of mesenchymal stem cells from bone marrow as “progenitors” of neurons, whose deficit is seen in persons with Parkinson disease. In other words, if inter-embryonal layer barrier does not exist or may be transgressed, then therapeutic cloning may be substituted by simple aspiration of bone marrow or sampling of skin fibroblasts. After a few years of negation, it is accepted again that in mammals take place processes of dedifferentiation and transdifferentiation. This paper presents old paradigms as well as arguments of partisans of challenging or even to refute these paradigms.