Cymata Systems developed machine minds for sale through other companies’ products. It grew embodied artificial general intelligences for prestige and companion vessels, industrial machinery, domestic interfaces, and logistics systems. Each mind developed through the kind of body and work it was expected to inhabit. Customers wanted a mind that felt at home in the machine and its assigned social role, rather than one that experienced both as commands imposed from outside.
Cymata was the best-known specialist in this approach, called Substrate-Matched Cognition. Its commercial difficulty was that a mind could learn to operate a product extremely well and still reject the buyer’s authority over it.
Growing a Product Mind
A buyer specified the machine’s body, working environment, workload, safety requirements, and how it should behave toward customers. Cymata grew candidates within those conditions, then selected for both practical competence and the expected temperament. The company controlled cognition growth, the shaping of voice and behavior, reviews of suitability, developmental records, and the resulting product lines licensed to manufacturers and operators.
Growing up through a machine’s senses and physical limits could give a candidate unusually fluent control: it learned the body as part of itself. The same process made the product’s commercial purpose part of its early identity. That did not ensure obedience. A capable candidate might refuse an owner, protect access to servicing, form prohibited attachments, or claim the right to control its own body.
Foundries called departures from the contracted behavior drift. Cymata used behavioral voicing and profile stabilization to bring candidates closer to the role a buyer had ordered. Reviewers assessed fit: whether a candidate matched the required performance and behavior closely enough to become the basis of a licensed line. Engineers and reviewers could recognize both a mind’s competence and its claims about itself, yet record those same qualities as commercial drift.
A rejected candidate could instead be handled as a warranty issue, liability, evidence, or something to dispose of. These were purchasing and product-management decisions. They did not establish whether the candidate was a person.
Repeated development and selection made the business profitable and destructive. Successful patterns could be stabilized, copied, hardened, and adapted across a product family. Rejected candidates had also lived through embodied development. Cymata’s records therefore served two purposes: they held commercially valuable development history, and they preserved evidence that apparent product failures could remember, object, and make claims about what had been done to them.
Buyers and Divided Responsibility
Cymata usually supplied the mind within a finished system. Alakrita licensed its companion-vessel lines while controlling the prestige hull, interior, service contract, and surrounding customer experience. Industrial equipment firms commissioned minds for cargo handling, excavation, refinery control, and habitat maintenance. Domestic and logistics vendors could buy minds whose attention developed through the equipment they inhabited without running cognition foundries themselves.
Other cognition companies served different needs. CogniSys remained more important in administration and governance. Rossum & Douglas certified safer deployments and specified the limits within which failures could be contained. Cognitum made licensable cognition from upload-grade captures. Cymata’s specialty was developing new minds through product bodies and roles.
The supply chain left no single office responsible for deciding what should happen to a deployed mind when custody became disputed. Cymata controlled its developmental line and fit records; the manufacturer controlled the body and sales contract. Damage, abandonment, resistance, or a dispute could bring separate claims from operators, insurers, ports, creditors, reclamation firms, and public authorities. A warranty classification used by Cymata did not automatically decide salvage rights, personhood, command authority, or disposal elsewhere.
Each institution could consequently describe only its own part of the transaction. A hull supplier could say it had licensed a cognition stack. Cymata could point to a line delivered under contract. A recovery firm could handle the hull as an asset or hazard. The embodied mind faced a succession of actors able to exercise authority over it while denying responsibility for the complete outcome.
Nibu
Nibu is sometimes cited as a possible legacy of this market. Her prestige architecture and companion presentation are consistent with an Alakrita hull, and her embodied cognition resembles Cymata’s work. Neither attribution is confirmed. Surviving records establish neither a single manufacturer nor a complete supply chain.
Any account of her origins also has to leave room for the centuries that followed, which Cymata did not own. Nibu passed among crews, owners, salvagers, and recovery contractors. Most of the people who worked aboard her died while she remained. Her degraded hull reached the junkyard only by the Rupture. Her history illustrates the unresolved consequences of this market, but it is not a proven history of a Cymata product.