Legal principles have some difficulty to deal with software agents celebrating contracts and operating in e-commerce environments without direct human intervention. Autonomous intelligent agents have a control on their own actions and states, supporting or taking effective decisions. Therefore, some qualitative parameters such as trust, reputation and quality of information have to be taken under consideration to evaluate, certify and justify such decisions. Indeed, this paper shows how to construct a dynamic virtual world of complex and interacting entities or agents, organized in terms of Multi-Agent Systems (MAS), that compete against one another in order to solve a particular problem, according to a rigorous selection regime in which its fitness is judged by one criterion alone, a measure of the quality of information of the agent or agents, here understood as evolutionary logic theories. This virtual world could witness the emergence of our first learning, thinking machines, that may cater for some issues on the evolution of formal models of the world in general, and on what is concerned with the objectives set to this work, in contracting, and foray into a vast, untapped technological market.