Project Transcendence  ·  A research question offered for critique

What must remain continuous for the same “I” to survive a change of substrate?

Not a personality, a memory archive or a convincing reconstruction — the continuation of this experiencer. We break the question into parts that can be tested, criticised and, where necessary, proven wrong.

Not a claim of discovery. No technology described here preserves, uploads or transfers a mind.

I

The question

Three propositions that must never be merged

Public talk about “uploading” runs three different questions together. Evidence for the first does not establish the second, and neither establishes the third.

P–I

The system preserves information about a person

Typical evidence
Fidelity of memories, preferences, speech and behaviour.
Cannot show
That the system is conscious, or that it is the original subject.
P–C

The system is conscious

Typical evidence
Theory-dependent markers: report, behaviour, neural or architectural indicators.
Cannot show
Whose consciousness it is, or whether it continues an earlier subject.
P–S

The original subject continues

Typical evidence
No validated direct measure exists.
Status
Unknown The heart of the project — and possibly permanently underdetermined.

A faithful copy could remember everything and sincerely report that “it worked” — and the third question would still be open.

Copying, or functional similarity, is not preservation of the original experiencer.

Three tracks, kept strictly apart

A

Inspiration

Philosophical, spiritual and artistic sources. They generate questions; they carry no evidential weight.

B

Speculative engineering

Pre-feasibility concepts with kill criteria. No role in any consciousness claim.

C

Testable science

Narrow questions with measurable falsifiers, open to outside criticism.

II

Eight continuities

What exactly is supposed to carry on?

“Continuity” means at least eight different things. Each comes with a boundary — the thing it does not, by itself, settle.

01

Biological

Persistence of the living organism and its relevant physiology.

Does not by itself settle what constitutes subjectivity.

02

Neural / dynamical

The ongoing neural organisation and causal dynamics, maintained through change.

Which dynamics are necessary — or sufficient — is open.

03

Psychological

Memories, dispositions, values, relationships, self-narrative.

A successor can preserve similarity.

04

Informational

Data sufficient to model, reconstruct or emulate aspects of a person.

Preserving information is not survival of the subject.

05

Causal

Later states arise from earlier states through one ongoing process.

A candidate identity condition, not a demonstrated one.

06

Subjective

The present experiencer is the subject of later experience.

The central unresolved issue.

07

Social

Influence, works, relationships and memories held by others.

A legacy can continue after its subject dies.

08

Spiritual

Survival or transformation of a soul or spirit.

Metaphysical · Track A No empirical mechanism is proposed.

III

Hypotheses & falsifiers

Eight claims, each with a way to prove us wrong

“First-person discontinuity” cannot be measured, so we test what can be: report, behaviour and neural indices of interrupted experience. None of these proxies can detect whether the original subject continued. Their job is to narrow the question honestly.

H1Conceptual · expert review

Identity theories give observably different verdicts across replacement scenarios.

What would prove us wrongTwo rounds of independent review find no divergent case with an agreed third-person discriminator. The question is then recorded as empirically underdetermined — a valid result.

H2Empirical · first priority

“Interruption of experience” — anaesthesia, dreamless sleep, meditative cessation — splits into separately measurable components.

What would prove us wrongA systematic audit finds report, behaviour and neural indices never dissociate, or none is reliable enough to use.

H3Methodological

People’s sense of being “still the same experiencer” sits near ceiling, so it discriminates poorly.

What would prove us wrongPre-registered measurement finds reliable between-condition variance in continuity judgements.

H4Empirical · audit

In closed-loop prostheses and BCIs, embodiment and agency can change independently of task performance.

What would prove us wrongNo dissociation is found, or no validated instrument exists. Integration is not extension of the subject.

H5Empirical · audit

Memory prostheses improve the recipient’s own encoding; they do not import someone else’s content.

What would prove us wrongA controlled study shows acquisition of specific, donor-unique content. Even then: memory transfer is not experiencer transfer.

H6Neurophenomenology

Altered-state self-model changes show reproducible structure, strongly shaped by context and expectation.

What would prove us wrongNo factor structure replicates beyond expectation and demand effects. Perceptions in altered states are not external events.

H7Agency · safety layer

Training metacognition improves confidence calibration and reduces compulsive behaviour over time.

What would prove us wrongNo advantage over an active control at the pre-registered endpoint. Ethics review first; says nothing about identity.

H8Biology · audit

Organisational pattern can persist through component turnover in regenerating tissue.

What would prove us wrongEffects prove transient, unreplicated or confounded — and the analogy is downgraded to metaphor. Morphogenesis is not a theory of consciousness.

IV

Research programme

Narrow next steps — plans, not results

We make theories commit to explicit, comparable predictions, then ask the people best placed to break them.

Prediction matrix v0.1 Is it the same subject afterwards? Our inferences, drafted so experts can correct them — not a consensus of the authors cited.
ViewGradual replacementScan-and-copyFissionAnaesthesiaPartial prosthesis
AnimalismS?NSNSS
Psychological continuitySSNBSS
Closest continuerSSBSS
Structural / patternSSBSS
Stream continuityS?N??S
Deflationary???SS
Can any observable bear on it?Detects alteration, not P–SNoNoPartly — the first test casePartly (H4)

S same subjectN not the same? indeterminateB branchingSwipe the table sideways →Two codes in a cell = the view splits on versions or conditions.

  1. Draft v0.1 · published

    Prediction matrix

    Identity and consciousness theories × five scenarios. Version 0.2 comes from two rounds of structured expert review, with agreement statistics per cell.

  2. Planned · first priority

    Interruption-of-experience audit

    Anaesthesia and recovery, NREM sleep with serial awakenings, meditative cessation. A states × measures table: report, behaviour, neural indices — with reliability, dissociations and confounds.

  3. Planned

    Prosthesis & memory-prosthesis audits

    Studies reporting both performance and validated embodiment or agency instruments; memory-prosthesis work classified as content-specific or facilitation.

  4. Planned · no one contacted yet

    Outreach for critique

    We will ask consciousness scientists, cessation researchers, anaesthesia groups and philosophers of identity for three things: the strongest objection, the most relevant existing work, and the smallest test that could falsify one part of the programme.

  5. Standing rule

    Ethics first

    No human data collection without independent ethics review, consent and stopping rules. No self-experimentation. Nothing here recommends or supplies any substance.

  6. Separate track · pre-feasibility

    PPAS — a self-repairing plate

    Engineering only: can a sensor-equipped plate detect, localise, repair and verify a small leak, benchmarked against simpler actuators and governed by kill criteria? It plays no role in any consciousness claim.

V

What we do not claim

The project does not claim, and has not demonstrated:

  1. Technological immortality, mind uploading or substrate-independent consciousness.
  2. Any sufficient test for survival of the same first-person subject.
  3. That any current technology — BCI, neural prostheses, whole-brain emulation, AI, regenerative medicine — solves continuity of consciousness.
  4. That an accurate AI replica is the original person.
  5. That transferring memories transfers their original experiencer.
  6. That reconstruction continues a particular earlier subject.
  7. That cryptography or distributed storage is a mechanism of soul or consciousness.
  8. That ancient or religious texts encode modern neurotechnology.
  9. That anything perceived in altered states occurred in external reality.
  10. That the engineering track is conscious, would preserve a person, or is meant to house one.
  11. That any engineering target has been achieved.
  12. That any plan on this page has been carried out.

The question may turn out to be permanently underdetermined. That would be a result, not a failure.

VI

Myth layer · Track A

The art may depict anything. The captions claim nothing.

Images keep the question alive in a way an argument cannot. Every image here is labelled as metaphor and carries no evidential weight.

Hieronymus Bosch, Ascent of the Blessed: angels carry souls upward toward a tunnel of concentric rings opening into light.
Metaphor · not a claim

Hieronymus Bosch, Ascent of the Blessed

c. 1500–1504 · oil on panel · from Visions of the Hereafter · Gallerie dell’Accademia, Venice · public domain

Souls are carried up through a tunnel of rings into light. Read only as an image: the passage is a process, not a jump — and the soul is carried through, not copied across.

Visionary illustration: a winged figure in a halo of gold light points toward a radiant emblem above a city of spires.
Metaphor · not a claim

Vision study, chapter i

Project artwork · Metanoeticist

The serialized story of Transcendence lives here, in pictures. It illustrates questions and never answers them.

Three images we return to Metaphor

The eggshell
What does a long-lived structure need if its material is constantly renewed?
The droplet and the fire
Ideas passing from one mind into others — social and informational continuity, not subjective.
Blood and fibrin
Seal now, rebuild later: the logic of repair, borrowed for engineering only.

Tell us the strongest objection.

We are looking for researchers willing to name the best counter-argument, the most relevant existing work, and the smallest test that could prove one part of this wrong.