A decade of work. A sequence of proof.

2026–2028: establish a verifiable archive

The first proposed milestone is operational, not speculative: define and validate the archive service. Candidate work includes a limited specimen panel, professional storage agreements, a digital accession system, consent records, and transfer procedures. Death X has not demonstrated these as an operating service.

An acceptable result would be a complete, inspectable record for each accepted archive: identity and custody documentation, material-specific quality results, recoverability evidence where appropriate, digital integrity records, and a tested export. A failure to recover a sample or open an archive is a finding to disclose and investigate.

2028–2031: test source-grounded digital representations

The next proposed workstream would evaluate whether a consented AI representation can answer questions using the person’s actual archive. Evaluation would separate quotations, supported summaries, uncertain recollections and invented responses. A model should be able to say that the record does not contain the answer.

Testing would use questions withheld from system development, source citation checks, family feedback where authorized, and challenges designed to expose false confidence. Synthetic speech and generated video would remain visibly labeled. The original files must remain accessible independently of whichever model is used.

2031–2036: evaluate robotic embodiment

A longer-term research objective is to connect a permissioned digital representation to a physical interface, potentially including a humanoid robot. Relevant questions include whether speech, gestures and retrieval remain faithful to source material, whether bystanders understand what the system is, and whether it can operate safely within a tightly defined role.

A meaningful demonstration would be a controlled interaction supported by consent, traceable responses, independent evaluation, and a clear shutdown procedure. A humanoid appearance or familiar voice would not demonstrate that the original person’s consciousness has continued. Those are different claims requiring different evidence.

Biological research runs on its own evidence track

Cell and organoid research could improve the usefulness of preserved biological material. Each proposed experiment would require a specific question, suitable oversight and consent, and a testable outcome. Research results must be described at the level actually demonstrated: cells, tissue models, or particular functions.

The program does not assume that a digital biography can be written into a donor-derived organoid, or that preserved cells plus an AI model become a continuing person. There is no demonstrated integration pathway supporting that claim. Such an assumption cannot be used to price or promise a future product.

The gate is evidence, not the calendar

These dates are planning horizons proposed in October 2026, not delivery commitments or forecasts of scientific breakthroughs. Funding, collaborators, facilities, legal review and experimental results will determine what proceeds. If a stage fails its criteria, the appropriate response is to revise or stop it.

Each public milestone should identify the method, comparison, measured outcome, failures, independent review and remaining limits. A publication from another laboratory is useful evidence for the field; it is not proof that Death X has reproduced the result or delivered a service.

What would count as progress for a member?

Early progress would mean better documented materials, more durable access and fewer unknowns about custody. Later progress could mean a more reliable way to search a life archive or interact with a permitted representation. Those outcomes can be valuable without claiming a transfer of subjective experience.

The strongest version of the program keeps its ambition and its accounting equally clear: what was preserved, what was measured, what was demonstrated, and which question remains unanswered.

Sources & further reading

1. ISSCR — Guidelines for Stem Cell Research and Clinical Translation ↗
2. National Cancer Institute — 2026 Best Practices for Biospecimen Resources ↗

Linked research supports the specific statements discussed. It does not validate the proposed Death X service. Sources checked 14 October 2026.