In this review will be underlined two simple ideas of potential interest for the design of cancer immunotherapies. One concerns the importance of kinetics, with the key notion that a single cause may trigger two opposite effects with different kinetics. The importance of this phenomenon will be underlined in neurobiology, transcription networks and the immune system. The second idea is that efficient immune responses have been selected against pathogens, throughout evolution. They are never due to a single cell type, but always require multiple, complex cellular cooperations. One cannot recognize this fact and persist in the presently dominant T-cell centered view of cancer immunotherapies. Suggestions will be made to incorporate these simple ideas for improving these therapies.