The theorem states that any projection made of a stationary Markovian dynamics by conditional probability will result in the induced image of that Markov chain to have increased entropy, due to the Markov property and conditioning reducing entropy. The use of a network of conscious agents as a probability space with the Markovian dynamics can aid in understanding this concept.
The discussion revolves around the interesting distinctions between the ideas of consciousness and the fundamental disagreement regarding space-time between the speakers. The speaker also mentions the technical term "conscious agent" and explores the connection between consciousness and self.
The entropy of a stationary Markovian dynamics is constant, meaning that there is no entropic time. However, any projection made of that dynamics by conditional probability will have increasing entropy, resulting in entropic time.
The Markovian dynamics is a way of modeling a system with probabilities and agents that interact in a specific way. Stationarity is one of the properties of this method that has certain properties in terms of entropy that scientists can take assumptions from.