Several problems, almost impossible to defeat, among them, heat removal, addressing small devices, and fuzziness at atomistic scales confronts standard CMOS electronics. What we are concluding after intensive research is that the material is not the major problem but the way how we encode information is what can allow us to continue the steady exponential grow in computational power know as the Moore's law. Although molecules and nanoclusters are the alternative for scaling down devices, we need to develop scenarios for information coding and transfer in molecular circuits able to operate at integration densities and speeds orders of magnitude higher than in present integrated circuits. A simple scaling down using the same scenarios being used in today's microelectronics devices does not offer much hope at the atomistic and nanoscopic levels. We proposed two scenarios: One is based on the characteristic vibrational behavior of molecules and clusters and the other is based on their molecular electrostatic potentials. It is proposed that these two scenarios can be used for molecular signal processing and transfer in molecular circuits; theoretical demonstrations using computational techniques are presented for these two paradigms. The molecular electrostatic potential in the neighborhood of a molecule has very well defined zones of positive and negative potential that can be manipulated to encode information and vibrational modes of long molecules can allow us to transfer signals. Both scenarios allow very lower energies, higher speeds, and higher integration densities than in any other technology. A review of our search for other scenarios for coding, processing and transport of information is provided.