Onboarding Immortality

cLife is an open research program for human–AI continuity.

We study non-invasive brain–computer interfaces, longitudinal identity signals, personal AI, and resilient computing. Our long-term continuity hypothesis is speculative; every step begins with measurable experiments, informed consent, reproducible methods, and clear limits.

Join the Beta Read the Manifesto

PUBLIC RESEARCH STATUS

cLife Research Program

Current scope:
  ✓ non-invasive EEG experiments
  ✓ public-dataset benchmarking
  ✓ longitudinal self-model studies
  ✓ consent and neural-data governance
  ✓ reproducible open-source pipelines

Research hypotheses:
  ◇ multimodal “echo fingerprints”
  ◇ human-guided personal AI continuity
  ◇ identity drift and correction metrics
  ◇ resilient, user-controlled archives

Not claimed:
  × consciousness upload
  × proven digital immortality
  × clinical diagnosis or treatment

BCI

Non-invasive Neural Interfaces

We begin with EEG and related biosignals to study attention, evoked responses, motor imagery, signal stability, and human-controlled interfaces.

DIGITAL SELF

Continuity as a Hypothesis

We test whether memories, values, language, decisions, and neural patterns can form a longitudinal model—without claiming that a model is the original person.

SECURITY

Cognitive Sovereignty

Participants should control access, reuse, deletion, export, and withdrawal. Neural and autobiographical data must never become an irreversible product.

ETHICS

Ethics Before Scale

Human-subject research requires independent review, informed consent, minimization of risk, privacy protection, auditability, and clearly stated research limits.

RESEARCH QUESTION

Can a digital echo remain meaningfully aligned with a living person?

We call the gradual preservation, transformation, and divergence of a person’s digital traces Echotropy. The program studies measurable continuity—not metaphysical proof of survival.

Neural continuity: stability and drift of non-invasive EEG features across days, months, tasks, and devices.
Narrative continuity: stability of values, memories, preferences, language, and self-description.
Behavioral continuity: agreement between a participant and a personal AI across decisions and scenarios.
Corrective continuity: whether participant feedback can reduce model drift over time.
Embodied continuity: how sleep, stress, movement, physiology, and context alter any apparent “self signature.”

RESEARCH BOUNDARIES

What the work can—and cannot—establish

EEG measures electrical activity at the scalp; it is not a complete reading of thoughts or consciousness.
A personal model may reproduce patterns while lacking subjective experience.
Re-identification accuracy is not proof of identity persistence.
Longitudinal similarity may reflect stable habits, context, physiology, or data leakage.
Every result must be compared against simple baselines, held-out sessions, and alternative explanations.
No clinical diagnosis, treatment, implant, or invasive procedure is part of the initial program.

ACCESSIBLE EEG

Rapid Prototyping

Low-channel wearable EEG such as Muse can support onboarding studies, resting-state recordings, basic spectral analysis, and early stimulus experiments. These systems prioritize speed and accessibility over spatial coverage.

Muse 2 specifications ↗

OPEN BCI

Flexible Research

OpenBCI Cyton provides eight channels and can be expanded to sixteen with Daisy. It is suitable for custom electrode placement, EEG/EMG/ECG acquisition, reproducible prototypes, and hardware-agnostic software development.

OpenBCI Cyton documentation ↗
Cyton and Daisy specifications ↗

HIGH-DENSITY

Advanced EEG & Multimodal

Research systems with 8–64 EEG channels, and optional EEG–fNIRS combinations, can support higher-density studies, multimodal validation, mobile experiments, and stronger comparisons across sensor classes.

g.Nautilus Research ↗
g.Nautilus EEG + fNIRS ↗

MULTIMODAL

Beyond EEG

Future non-invasive studies may combine EEG with eye tracking, ECG, EMG, respiration, motion, voice, behavioral tasks, sleep measures, and fNIRS. Each modality should answer a defined question—not simply increase data volume.

MNE neurophysiology tools ↗

Transparent Roadmap

From reproducible signals to continuity research

Stages are evidence gates, not promised dates. We advance only when methods, ethics, and replication are strong enough.

FOUNDATION

Governance, protocols, and public baselines

Publish the research charter, consent model, data-retention rules, threat model, exclusion criteria, and definitions of “continuity,” “drift,” and “digital echo.” Reproduce established analyses on public EEG datasets before collecting private neural data.

Outputs: protocol registry · ethics checklist · BIDS data template · public baseline notebooks · risk register

SIGNAL PIPELINE

Device-agnostic acquisition and quality control

Build synchronized acquisition using BrainFlow and/or Lab Streaming Layer, then process with MNE-Python. Validate timing, dropped packets, electrode quality, artifact handling, reproducibility, and data export across accessible and research-grade devices.

Outputs: acquisition SDK · live quality dashboard · preprocessing pipeline · device comparison reports

CLASSIC BCI

Replicate established paradigms

Run resting-state, eyes-open/eyes-closed, SSVEP, P300, motor imagery, and attention tasks. Pre-register metrics and evaluate within-session, across-session, and across-participant generalization.

Outputs: replication reports · benchmark datasets · latency and accuracy results · failure analyses

LONGITUDINAL

Echo Fingerprint study

Collect repeated non-invasive recordings alongside structured interviews, preferences, values, and behavioral tasks. Test which features remain stable, which drift, and whether models generalize to future sessions without leaking identity through trivial metadata.

Outputs: continuity index proposal · drift curves · cross-day verification · confound analysis

PERSONAL AI

Participant-controlled Echo Agent

Create a local or self-custodied personal model from explicitly approved material. Compare its answers with the participant over time, show uncertainty, record provenance, and make every memory editable, exportable, and deletable.

Outputs: personal data vault · provenance viewer · answer-alignment studies · deletion and export tests

CORRECTIVE LOOP

Test whether drift can be reduced

Allow participants to inspect, dispute, and correct their echo. Measure whether active correction improves future alignment or merely overfits to recent feedback. Compare static archives, continuously trained systems, and retrieval-based models.

Outputs: correction protocol · drift audit · model lineage map · reversible update system

RESILIENCE

Long-term preservation without identity claims

Research encrypted replication, checksums, versioned memories, format migration, hardware failure recovery, offline operation, and inheritance controls. This stage preserves information and agency; it does not prove transfer of consciousness.

Outputs: continuity archive spec · recovery drills · cryptographic integrity reports · succession controls

INDEPENDENT REVIEW

Replication, external ethics, and publication

Invite independent laboratories, neuroscientists, philosophers, security researchers, and participant representatives to reproduce results and challenge the assumptions. Publish negative findings and abandoned hypotheses.

Outputs: external replications · red-team reports · peer-reviewed papers · public changelog

Open Research Stack

Tools, standards, datasets, and ethical references

Acquisition

BrainFlow — device-agnostic biosignal API.
Lab Streaming Layer — synchronized multimodal streams.
MNE-LSL — real-time integration with MNE.

Analysis

MNE-Python — EEG, MEG, ECoG, sEEG, and fNIRS analysis.
Braindecode — deep-learning workflows for EEG.
MOABB — reproducible BCI benchmarking.

Data Standards

BIDS — neuroimaging data organization and validation.
EEG-BIDS — EEG-specific structure and metadata.
HED — structured event annotation.

Public Data

OpenNeuro EEG — public BIDS datasets.
NEMAR — electrophysiology data and tools.
PhysioNet — physiological signal datasets.

Hardware

Muse 2
OpenBCI Documentation
g.Nautilus Research
g.Nautilus EEG–fNIRS

Ethics & Rights

UNESCO Recommendation on Neurotechnology Ethics
Declaration of Helsinki
Belmont Report

Transparency commitment: cLife will distinguish demonstrated results, preliminary observations, engineering targets, and philosophical hypotheses. Research involving people will not begin without an appropriate ethics pathway, informed consent, privacy controls, and a defined withdrawal process.