The state of the art of the modern production of cast iron, ferroalloys, and steel is analyzed. The existing two-stage scheme, which resulted from long-term improvement of the old well-known technologies, is shown to have a low susceptibility to new scientific achievements and an insignificant reserve of further improvement. The necessity of developing fundamentally new technologies based on the achievements of modern science is grounded, and this necessity is especially important for the stage of metal extraction from ores. As an example, the possibility of selective reduction and extraction of metals from ores without melting is shown.