Keep a record
Files, sequence and provenance.
What can persist from a person—and what would count as continuation? These research directions are a map for investigation, not a menu of available procedures.
The vision of a person becoming something brain-like raises a problem that a change of material does not solve: how would the information supporting that person’s memories and identity survive? The research linked here does not show that dissolving a body creates a brain or carries a person into one.
A more useful research program separates what is retained—records, DNA, viable cells, tissue structure or function—from what can be built later. The connection to personal continuity remains an open question.
Files, sequence and provenance.
Cells, tissue and measured function.
No demonstration identified in this review.
Preserve original media, context, consent and exportability. This is the most directly specifiable near-term work.
A complete test archive must be exported, verified and restored by a second custodian.
File access cannot continue a person’s subjective experience.
Retain the measurements and uncertainty behind a genome interpretation, including future reanalysis needs.
Specify raw data, assembly, quality thresholds, access controls and independent identity checks.
A genome does not contain a recorded life or a complete learned brain state.
Explore whether banking viable donor cells adds research options beyond banking sequence data.
Define intended uses, post-storage quality tests, consent boundaries and qualified custodians.
Cell survival does not establish personal survival.
Study how cells can form and sustain neural tissue with measurable properties.
Characterize cell types, variability, function and oversight before any individual-continuation interpretation.
New donor-derived tissue does not inherit the donor’s autobiographical memories.
Investigate the scientific and operational demands of maintaining neural models over years.
Track function, stability, contamination risk, failure response and governance across time.
Five years of developmental maturation is not five years of a person’s remembered life.
Compare what living neural networks can do when connected to inputs, outputs and computational readouts.
Use repeatable tasks, controls and whole-system resource accounting.
Learning a task does not establish consciousness, identity or a recovered mind.
Better measurements may distinguish meaningful changes in tissue rather than relying on appearance.
Establish recording reliability, signal interpretation and validation across independent models.
More electrodes do not themselves decode a life.
Use public circuit maps to study what structure can explain about function.
Reproduce a bounded analysis, preserve the dataset version and test a specific prediction.
A wiring map is incomplete evidence for the full state of a mind.
Follow work that evaluates function after preservation, alongside structural and molecular endpoints.
Separate viability, tissue function, duration, scale and species in every comparison.
An animal organ or neural-tissue result does not establish human revival.
Study which cellular processes remain recoverable under defined experimental conditions.
Map exactly which functions were recovered and which remained absent or untested.
Cellular recovery does not prove restoration of a coordinated brain or a person.
Compare encoded digital storage in molecular materials and living populations.
Measure payload recovery, drift, redundancy, custody and realistic costs.
A file encoded in DNA is not a biological recollection.
Investigate what adult-derived cells can form and how collective behaviour emerges.
Define the useful function and reproduce it with appropriate controls.
Movement and self-organization do not make a construct a brain or the donor.
Assess a small, clearly defined culture carrying a consented sequence and documented provenance.
Independent feasibility, genetic stability, containment, maintenance and applicable authorization.
No complete-body conversion or memory-transfer process is established here.
Keep original recordings separate from generated responses and give people control over use.
Document permissions, provenance, error disclosure, withdrawal and successor decisions.
An AI imitation is not evidence that the original person continues.
Fourteen directions are mapped here. This is a curated scoping review, not an exhaustive inventory of every publication, patent, company or experimental approach. Neural-model work would require appropriate donor consent, qualified scientific oversight and attention to emerging welfare questions. No human dissolution or revival procedure is offered.