Build the vessel.
Preserve the person.
Building the foundations for persistent digital identity—and investigating the much harder question of cognitive continuity beyond a biological body.
Research status documented September 25, 2026. A working prototype and research results are not yet verified.
The long-term ambition is to investigate technological immortality: whether an individual’s cognition could be preserved and potentially continued across computational systems and embodiments.
The first step is concrete. We want to build a digital persona with a traceable history, a persistent virtual space, and defined software state that can be saved, restored, and tested.
A convincing recreation would not establish that a person’s experience continued. Orpheus keeps that distinction central to the work.
Plan persistent identity, source-traceable memory, a virtual habitat, and interfaces for voice, avatars, and compatible devices. Start with an original synthetic persona; treat any representation of a real person as a separate, consent-based undertaking.
Define the first software demonstrator and its acceptance tests.
Study published biological neural models through permitted datasets and simulations. Future work on human neural control and physical embodiment would depend on qualified partners, appropriate oversight, and demonstrated safety.
Select a published model result for a bounded reproducibility study.
Develop tests that distinguish stored records, preserved functions, individual fidelity, and human agency. The deeper question of subjective continuity remains open; negative and ambiguous findings belong in the record.
Define what each test could establish—and what it could not.
A focused software demonstrator. These are proposed development targets, with final scope and delivery still to be defined.
Versioned identity records, permissions, source citations, and clear handling of unknown or conflicting information.
Check retained records against a held-out test set and report errors.
One virtual room with saved objects, positions, and conversation context, plus a supported voice or avatar interface.
Show controlled changes surviving shutdown and restoration.
An authorized session accessed through two interfaces, with declared application state restorable on a compatible host.
Publish the state boundary, version manifest, and restore comparisons.
Setup instructions, test conditions, limitations, and an honest account of the demonstration.
Let an independent reviewer repeat the supported process.
An original test persona receives a permitted record, answers with its source, and moves an object in a virtual room. After shutdown and restoration, the defined record and room state are checked. A second interface then accesses the same authorized session.
This would test specified software functions under stated conditions. It would not demonstrate a human mind upload, biological-memory extraction, or preserved subjective experience.
The current foundation is a research thesis and drafted roadmap. The next milestone is a defined, testable prototype scope.
The project’s website now reflects the Orpheus name and visual identity. The research mission and proposed demonstrator remain the foundation of the work.
The version 1.0 project brief records the adopted thesis and proposes a demonstrator built around traceable memory, a persistent habitat, and verifiable state restoration. This is a planning milestone; technical implementation remains unverified.
Build an embodiment-independent architecture capable of housing a persistent digital identity and progressively integrating neural control, sensory embodiment, memory, cognition, and external computation. Use increasingly capable biological interfaces to investigate whether human agency and cognitive function can be continuously distributed from biological tissue into an artificial substrate, ultimately studying whether continuity can persist through the natural loss of biological function.
This is a research objective, not an achieved result or a promise of future capability.
Do the declared records and software state survive a restart or compatible restoration?
Are defined capabilities preserved under controlled tests, with errors and changes measured?
Does the same individual’s experience continue? Stored records or convincing behavior alone do not establish this.
Orpheus is not offering mind uploading, medical treatment, or enrollment in human experiments. Future human or end-of-life research would require qualified partners and independent clinical and ethical oversight. Following or supporting the project does not confer access to such research.
Published work provides questions and methods to investigate. These are external studies, not Orpheus results or endorsements.
A structural map of an adult fruit-fly brain, relevant to studying neural connectivity.
A computational model examining circuit properties involved in feeding and grooming.
A study of brain-to-voice synthesis, relevant to a defined neural communication capability.
Founder & Chief Developer, CMCAI
Connor leads Project Orpheus with a commitment to making the research question, proposed work, and limits understandable. The aim is to document design decisions, real tests, and setbacks as the work develops.
Orpheus has its own research scope. Central AI remains CMCAI’s workforce and governance system; Nexus and Norse Build are separate projects.
Meet the founderFollow CMCAI’s channels for project news. Builders, researchers, and potential sponsors can contact Connor to discuss a defined contribution or next milestone.
Persistent software state. Memory provenance. Interfaces. Computational neuroscience. Evaluation and consent.
Discuss Project OrpheusFor project conversations only. Please do not send medical, neural, or private autobiographical records.