Mind Architecture
The idea that prediction and reward are tools, but not by themselves a mind.
Mind architecture is the name for the recurring question behind Sapiexo: what kind of machine would need to exist if prediction, reward and scale are not enough?
Language models predict text. Reinforcement learning chases reward. Both are useful tools. Neither property, by itself, gives a system causal memory or self-existing continuity.
The public essays approach the same problem from different sides: proof, memory, causality, energy, development and the limits of scaling.