The definition of the right goal constitutes an essential part in evolutionary algorithms. The best EA is no use if the defined and calculated purpose does not correspond to a real need. Human needs are generally complex. That is the reason why users of EAs should be interested in utility theory and psychological mechanisms, up to the core of the single neuron. In this paper we show that the harmonic decision matrix opens a new way to model human decisions and basically, neuronal computation, that goes beyond the weighted sum and the polynomials.