During a gradual increase of stress on rock samples and before their rupture, electromagnetic anomalies have been recorded from DC-ULF, VLF, up to VHF frequency bands. On the other hand, broad range electromagnetic precursors falling in the same frequency bands have been internationally reported before large earthquakes. Thus, the technique of measuring electromagnetic radiation emitted during rock fracturing is a candidate for forecasting global failure both in the laboratory and the geophysical scales.Until now, the physical mechanism of the generation of these precursors remains an open book. An essential requirement for this type of knowledge is a careful laboratory investigation of rock failure. A similar investigation between the pre-fracture electromagnetic emissions in laboratory-scale experiments and tectonic-scale events is suggested. Accumulated laboratory evidence also suggests that both electromagnetic and acoustic emissions ("laboratory earthquakes") are two sides of the same coin; both are caused by the opening cracks population. Recently, electromagnetic and acoustic studies in terms of self-organized complexity indicate that the same dynamics may govern the large earthquake and the laboratory scale sample rheological structure. The above statements encourage the investigation of the connection between seismic, geodynamic, geodetic and electromagnetic investigations, as well as the dependence of this connection on the general properties of the under deformation complex system. Beyond these investigations, a number of other approaches (e.g. Thermal Infrared Radiances) observed both in the laboratory and the geophysical scales strengthen the idea that such multidisciplinary efforts will enhance the understanding of the physics behind these observations and thus take us one stop closer to fracture prediction in disordered media. The achievement of converging estimations would definitely improve the chances for an earthquake prediction. The earthquakes are made of many highly interconnected parts on many scales. The improvement of their prediction must be the result of many highly interconnected scientific teams.