Research this ambitious needs to move slower than the hype around it.
MemClone exists to build the underlying architecture for reading, storing, and — eventually, safely — continuing human memory. We think that work deserves the same rigor, transparency, and caution as any other frontier of medicine.
Our Mission
Every technology that has ever extended what a human mind could do — writing, photography, recorded sound — began as a way of getting a memory outside the body that made it. MemClone's mission is to build the scientific and engineering foundation for the next step in that line: reading the brain's own engram structure directly, with enough fidelity that a life's memories, voice, and personality could someday be preserved as more than an archive — as something a person's family could still recognize and, if they choose, still speak with.
We are a research organization first. We do not sell a working memory-cloning product today, because one does not exist yet — anywhere. What we are building is the architecture, the hardware roadmap, and the ethical framework that this work will need once the science catches up to the ambition.
Our Approach
We split the problem into what we believe is solvable this decade and what is not. Reading — high-density cortical and subcortical recording, edge neuromorphic compression, and long-term structural mapping of engrams — is an extraordinarily hard but tractable hardware and machine-learning problem, and it's where we focus our own research. Writing — recreating a memory inside a second brain — is, based on everything publicly known about neural stimulation safety, not achievable without unacceptable risk today, and we say so plainly rather than implying otherwise.
Where large language models come in is downstream of the hardware: engram data, combined with a person's own writing, speech, photos, and video — always with their explicit, revocable consent — can train a model that carries their voice, their reasoning style, and their memories forward as a digital presence, long before whole-brain read-write is solved.
Scientific Rigor
We publish our architecture for scrutiny rather than marketing it as finished technology. Claims are labeled theoretical until they're validated.
Consent & Human Dignity
No engram data, likeness, or personality model is ever built from a person without their own explicit, informed, and revocable consent — including for research on donated tissue.
Radical Transparency
We're explicit about what's proven, what's theoretical, and what's simply not possible yet — starting with the read/write asymmetry at the center of this whole field.
Long-Term Safety
Hardware that touches the brain gets held to the highest safety bar we know of, and we won't ship a "write" capability until the risk of tissue damage or seizure is solved, not just reduced.
Digital Personality Reconstruction
Long before whole-brain reading is possible, a meaningful amount of a person's memory, voice, and reasoning already exists outside their body — in writing, recorded conversation, photos, and video. Our personality-reconstruction research trains a personal language model on this material, combined with whatever engram data is available, to build a digital presence that a person's family can recognize as continuous with them: same turns of phrase, same recall of shared history, same way of reasoning through a problem.
This is opt-in work, built with the person themselves while they're able to participate, or with their estate's explicit authorization — never built from a person's data without that authorization.
Longevity & Digital Continuity
Our long-term research question is whether a sufficiently rich model of a person — memory structure, personality, voice, and reasoning, kept current through their life and beyond it — can constitute a meaningful, continuous digital presence rather than a static recording. We don't think that question has an answer yet, and we're not going to claim otherwise. What we can say is that the underlying engram-mapping and personality-modeling research is worth doing carefully, in the open, now — rather than waiting for someone to do it carelessly later.
Have questions about our approach?
Researchers, ethicists, and potential collaborators are welcome to reach out.