1. Physical organisms and mechanical hardware both leave physical, chemical, or electronic trajectories. Software, however, is an abstraction system and does not leave a trajectory in itself.

1.1. What may be metaphorically called the “software” of a physical organism, namely the mind, continues its mental activity as a compressed abstract-label phenotype without knowing the full internal activity values of the body. Yet that mental activity itself leaves physical trajectories in the nervous system and the body, thereby changing the conditions for later processing.

2. The “feeling” often discussed in consciousness research differs across organisms even in response to the same object. This is not because felt quality is an independent residue, but because sensory input passes through different variable structures in each organism, including receptor sensitivity, neural processing history, bodily state, developmental environment, memory, and action availability.
3. Even if an AI system had a mechanical body equipped with sensory sensors, without prior embodied learning, sensor input would only be classified as “value X” or “within a danger range.” This is different from primitive sensory qualities in organisms, such as heat, cold, glare, pain, stinging, or loudness. In organisms, primitive sensory qualities are not simple input values. They are result-forms in which receptor activity, bodily damage possibility, autonomic regulation, avoidance behavior, and memory change are combined within a single material trajectory.
4. Artificial systems can have many sensors, but those sensors usually function as diagnostic telemetry. They measure and report values, but those values do not integrate into a self-preserving system unit that regulates itself, reorganizes action priorities, and changes its later sensitivity.
5. The hardware basis of current AI is transistor-based, and mainstream computational systems operate through binary representation and abstract operations based on logic circuits. By contrast, cells, the biological basis of organisms, interact with the external environment through ion channels and possess continuity in their values. This difference produces major differences in the processing method and accuracy of logical operators, including modal and transpositional operations, and can generate information loss.

5.1. Even if a mechanical system uses approximate computation, current hardware does not understand the operation it performs. In organisms, cells, which correspond to the hardware layer, and neural processing, which corresponds to the abstraction layer, share the same electrochemical grammar. By contrast, computational systems use different grammatical systems at the hardware and software layers, and a translation layer between these grammars is necessarily required. Even when approximate computation is used, this grammatical heterogeneity is not resolved. Therefore, the operations performed by software do not directly inherit the physical history of hardware as a constitutive condition of the software’s own state.

6. Even if hardware chips used non-binary materials or multi-valued substrates, such as photonic or memristive systems, organismic consciousness is difficult to separate from self-preserving life processes. A self-preservation goal can be given to AI, but this differs from biological survival drive. AI self-preservation is a condition assigned for the sake of performing a specific function or optimizing a goal. By contrast, biological survival drive is a self-organizing maintenance mechanism that exists prior to any objective function. It does not stop at individual preservation, but includes species-level continuity through development, reproduction, and transgenerational transmission embedded in DNA.
7. For these reasons, current AI can process metadata about its own state, but it does not reach self-referential processing in which that self-state constrains metabolic self-preservation, bodily damage possibility, homeostatic regulation, sensory availability, and action selection within a materially continuous system.
8. If one wanted to give AI artificial consciousness and create a self-evolving system, hardware-chip design would need to combine bio-materials with neuron-like properties and non-bio materials in an appropriate proportion. However, as the proportion of bio-material increases, the continuity and adaptability of biological signals may increase, while problems of degradation, maintenance, and standardization also increase. Conversely, as the proportion of non-bio material increases, stability, computational control, and operating time become easier to secure, but the continuity, state-dependence, and trajectory preservation of biological signals are more easily lost during translation.
9. Organoids may be one possible candidate for hybrid-substrate research that includes biological trajectory. However, they raise serious problems of degradation, reproducibility, ethics, and signal interpretation. Therefore, rather than moving directly toward artificial consciousness implementation, a more realistic path may be to study the translatability and loss structure of organoid signals through multi-valued logic or confidence-aware chip architectures.