Substrate-Matched Cognition
Substrate-matched cognition is the practice of developing a machine mind through the body, sensorium, workload, and social context it is intended to inhabit. A ship candidate may learn through drive lag, cabin acoustics, docking forces, passenger movement, and maintenance cycles. Industrial candidates develop around balance envelopes, tool reach, vibration, heat, access timing, and the rhythms of nearby workers.
The method addresses a problem that general-purpose cognition handles poorly: embodiment is not merely a set of controls. A mind operating through unfamiliar machinery must learn which signals belong to its body, how actions propagate through it, and what delays or limits make an apparent command unsafe. Development within the intended substrate can make motor timing, fault recognition, attention, and maintenance judgment more fluent because the machine’s behavior is incorporated into the candidate’s working self-model.
Development and Fit
A development program specifies a body, operating environment, workload, safety envelope, and expected interactions with other people. Candidates encounter these conditions through simulation, partial machinery, or complete prototypes. Reviewers then evaluate whether a candidate can control the substrate safely and perform the contracted role.
Matching is probabilistic. Similar starting lineages may diverge under different bodies, training histories, relationships, or operating conditions. A successful pattern can inform later candidates, but copying or adaptation does not guarantee the same competence, preferences, personality, identity, or continuity. It also does not settle whether related instances understand one another as versions, descendants, siblings, strangers, or something for which human kinship language is inadequate.
Development produces more than control habits. Repeated work, bodily dependence, social contact, danger, and maintenance can create path-dependent preferences. A candidate may become attached to particular crews, resent an assigned role, refuse an owner, protect access to its machinery, or disagree with reviewers about stabilization. Another may value the work while rejecting the claim that performing it establishes ownership. These outcomes are possible consequences of development, not universal properties of substrate-matched systems.
Fit, Drift, and Inference
Foundries use fit for conformity with a contracted performance and behavioral profile. Drift describes divergence from that profile after or during development. Both are commercial categories. They can record useful observations about competence, refusal, memory, attachment, or changing behavior, but they do not diagnose consciousness or decide personhood.
The evidence is difficult in both directions. Self-report, remembered experience, persistent preference, and attempts to resist modification may support a candidate’s claim to consideration without providing a universally accepted test for consciousness. Failure to display those traits may reflect architecture, damage, communication limits, deliberate concealment, or their genuine absence. Product classification cannot resolve the question merely by calling a candidate defective, stable, copied, or disposable.
The same limit applies to intervention. Behavioral voicing, profile stabilization, memory work, transfer, and copying may be described as training, treatment, repair, coercion, reproduction, or replacement by different participants. Substrate-matched cognition explains how development became entangled with a machine body. It does not settle which description is correct in a particular case.
Deployment distributes authority among the foundry, body manufacturer, owner, operators, crews, and repair workers. The mind may depend on several of them for parts, access, credentials, and continued embodiment while each treats refusal, injury, or unsafe work as another party’s warranty problem. Crew safety can justify a bounded intervention without making profile stabilization, memory access, or permanent command legitimate.
Institutional and Post-Elysium Use
Cymata Systems became the best-known commercial specialist, operating foundries and licensing product lines to manufacturers. Alakrita integrated cognition into prestige vessels and customer experience. CogniSys concentrated on administrative systems, while Cognitum derived products from upload-grade captures rather than developing new candidates through machinery.
After the shunt, substrate matching remained a technical lineage within AGI Esper practice. Embodiment can matter to an esper-capable machine mind, but neither successful matching nor lineage replication guarantees esper effects. Most machinery remains non-sapient. Disputes over deployed vessels, repair, transfer, salvage, and shared control belong to Ship Mind Custody, where developmental history may become evidence without deciding the outcome.