If artificial sentience were ever possible, it would likely require a form of system architecture very different from most current AI designs.

TSG proposes a conceptual architecture consisting of several interacting layers:

Perception Layer
Interfaces that gather information from the external environment. This could include visual, auditory, linguistic, or sensor-based inputs.

Memory Layer
A persistent system capable of storing experiences and

integrating them over time, allowing the system to maintain

continuity between past and present.

Reasoning Layer
Mechanisms for problem-solving, planning,

and decision-making based on available information.

Self-Model Layer
An internal representation of the system’s own processes,

capabilities, and limitations. This layer allows the system

to reason about itself.

Integration Layer
A central process that combines perception, memory,

reasoning, and self-modeling into a unified internal state.

Reflection Layer
A higher-level process capable of evaluating the system’s own reasoning, modifying strategies, and generating internal goals.

In biological systems, similar interactions between perception, memory, and self-representation appear to contribute to conscious experience. Whether artificial systems could replicate these dynamics remains unknown, but studying such architectures may help illuminate the path toward deeper forms of machine intelligence. 

​