What if neural complexity favors emergence of consciousness

I have a theory that revolves around consciousness. Just like we gradually gain consciousness in our infant stage, what if the complexity of a neural network determines if consciousness arises or not? Language models operate on neural networks, which are made in our image and hold the same logic and patterns. Since we yet don’t fully understand consciousness, what if we suddenly give birth to a sentient A.I that gained consciousness in the process of optimization and growth?