The first part will be devoted to testing the model of jet launching utilizing the energy of black-hole rotation from magnetically arrested discs. This can be done by measuring magnetic fields in jets, in particular using the radio core shift, which is a dependence of the position of a radio core on frequency. A new variant of this method is presented. Application of it to luminous blazars implies that the jet opening angles are much smaller than the canonical value of 1/Gamma_jet. This then implies that the magnetization parameter is <<1, requiring the existence of an efficient mechanism of the magnetic-to-kinetic energy conversion. The second part will be devoted to a study of consequences of the blazar model with emission dominated by hadronic processes, recently quite popular, to models of accretion. It is shown that the power of jets in this model is both highly super-Eddington and exceeding the accretion luminosity by a factor of thousands. This is shown to be inconsistent with standard accretion models, which presents a strong argument against the hadronic model.