From cells, to flocks, to brains, to human societies, to artificial superintelligence (ASI), cognition can be understood in terms of self-maintaining information flows through time and space. Information’s substrate-independence and interoperability made possible for symbolic representations such as the genetic code to parasite the laws of physics. These patterns progressively became more complex, going through major transitions which gave rise to large varieties of self-correcting patterns. The key to understanding the emergence and dynamics of these transitions can be found in the way spatiotemporal entities process and exchange information between each other, that is, the causal structure of their sensorimotor loops & information trading patterns. Replicating these effects artificially within computers – either recent von Neumann architectures or more unconventional computing – enables us to factor out the fundamental principles of life in the universe, & represent formally helpful concepts such as autonomy, complexity, cooperation & integration. In this talk, I will present an exploration how a theory of information dynamics, coupled with artificial life simulation techniques, can help us characterize the fundamental laws that give rise to collective intelligence in the universe.