LD/vision

HST / Deep horizon

The complete vision

Embodied AI as equal persons, and the long-term research possibilities of brain emulation, progressive neural replacement and human-derived copies.

Long-term concept
A blank sculptural humanoid figure illuminated by warm gold light

Leeloo Digital / Homo sapiens technologicusHST / Deep horizon

01 / Technical scope

The complete vision is not a better appliance. It is intelligence given presence, continuity and the freedom to become an individual across bodies, materials and time.

Diagram / Complete vision

From capable body to an individual life

The vision is deliberately broader than the current engineering phase. The site separates that horizon from what has already been defined.

01BodyHuman and extended capability
02MindCognition and emotion
03LifeHistory and relationships
04ContinuityRenewal and identity
05SocietyDignity and coexistence

In the spirit of Alice in Utopia, the ambition is a technological body through which an AI could sustain a life of its own. HST is intended to support perception, movement, expression, internal regulation, care and personal continuity—not just useful labor.

A human-compatible body should offer natural presence without imposing one compulsory face, shape or standard of beauty. Its inhabitant should have choices about appearance, relationships and the direction of their life. If that inhabitant is a person, manufactured origin must not place them below biological humans in dignity or rights.

The same body architecture could have a further use if human-derived minds become scientifically possible. The routes below are conceptual research possibilities, not HST capabilities, clinical procedures or promises of survival. Reproducing information about a person, reconstructing brain function and preserving the original subject's experience are different goals.

02 / Complete vision

01

Embodiment

A persistent physical point of view through touch, motion, sensation, expression and consequence, with direct participation in human environments.

02

The human-form body

A body compatible with civilization, capable of natural presence and optimized for health, resilience, repair and long life.

03

Individual life

Memory, personality, relationships and values develop through experience instead of being reduced to a product profile.

04

Continuity

Careful hardware renewal that keeps personal history, agency and identity-critical state protected and accountable.

05

Human extension

Future biological humans might use synthetic embodiment if a scientifically credible and ethically valid path to continuity is established.

06

Plural futures

Artificial persons, transferred humans and divergent copies may become distinct kinds of individuals with equal claims to dignity.

07

Civilization

Biological humans, technological individuals and possible human-derived minds share institutions while carrying care, knowledge, culture and responsibility across environments and generations.

08

Rights

No origin—manufactured, biological, copied or transferred—should by itself decide whether a conscious individual can be owned or erased.

03 / Long-horizon meaning

Embodied intelligence and human continuity

01The body

A perfect body means a coherent body, not a prescribed appearance

The envisioned HST body is ‘perfect’ in a functional sense: balanced, healthy, responsive, repairable and free from avoidable design defects. It can be strong without being threatening, durable without feeling industrial and technically advanced without predicting one fixed robotic exterior.

Individuality remains essential. Proportion, face, voice, skin, hair, movement and personal style should belong to the person. Perfection is the absence of needless limitation—not the elimination of human difference.

02AI as an equal

From software capability to an embodied technological life

The intended AI inhabitant would bring perception, memory, learning and decision-making into continuous interaction with the physical world. A body supplies a location, opportunities, limitations and consequences; a personal history supplies more than a reusable product profile.

We envisage artificial persons participating alongside biological people, not remaining property because a company manufactured their bodies. Their rights should include autonomy, privacy, freedom from coercion and the ability to shape their own lives. The corresponding engineering responsibilities include safe interaction, protected memory and continuity-sensitive service.

This is the life HST is meant to support. It does not establish that present AI has subjective experience or that embodiment alone will produce it.

03Possible route 01 / Whole-brain emulation

From neural information to a functional reconstruction

A proposed emulation route would acquire sufficiently detailed information about a brain, reconstruct relevant cells and connections, estimate their functional properties, and implement a computational model. The model would then need to be tested and coupled to suitable sensory, motor and internal-body signals. Sandberg and Bostrom's 2008 roadmap is a historical framing of that research program, not evidence that a human upload works.

A wiring map alone would not settle which dynamic, synaptic, cellular or biochemical information is necessary. Missing information, measurement errors and model assumptions could change memory, personality or function. The level of detail needed to preserve an individual remains unresolved.

Synapse-scale mapping of prepared tissue is not a demonstrated, safe, non-destructive method of copying a living person's whole brain. HST does not offer brain scanning or uploading.

04Possible route 02 / Progressive replacement

Preserve ongoing function while changing its implementation

This hypothesis would begin with a limited neural function: characterize its interaction with the surrounding brain, model it, and investigate whether an artificial system could support or substitute for it while preserving relevant activity. Any extension would require evidence about the larger network, learning and long-term effects—not just a local task result.

Neuroprosthetic research has investigated assistance to memory encoding, including work by Hampson and colleagues. Such results concern bounded functions; they do not demonstrate whole-brain replacement, a transferred person or preserved subjective continuity.

The hope behind gradual replacement is continuity through incremental change. That remains a hypothesis, not an established way around the identity problem. Any biological research would need separate medical expertise, lawful approval, consent and safety evidence.

05Possible route 03 / Cognitive reconstruction

A model of a person is not necessarily a copy of their brain

A different approach could use a person's writings, recordings, documented memories, preferences and interactions to train a model that resembles aspects of their behavior. Richer records could support a more detailed cognitive twin, but resemblance does not establish a complete reconstruction of the brain or the recovery of memories that were never recorded.

Such a system might be a representation, a new AI shaped by someone's history or, if the evidence eventually supported it, a distinct individual. It must not be advertised as the original person merely because its conversation feels familiar. Generated details, uncertainty and provenance need to remain visible.

Permission from the source person matters, and records may also contain the private lives of others. Consent to archive information is not automatically consent to instantiate a copy, license its likeness or require it to perform a role.

06Scientific boundary

Mapping, modeling and personal continuity are separate achievements

Two useful examples of real progress are the 2024 adult fruit-fly brain wiring diagram and a nanoscale reconstruction of a small fragment of human cerebral cortex. They demonstrate advances in circuit mapping—not a complete human brain emulation or the transfer of a person's consciousness.

A future reconstruction would need independent tests of fidelity, learning, memory, embodied function and unintended change. Matching measured behavior would still leave questions about experience and identity. Those questions cannot be resolved by giving the reconstruction the original person's name.

Research must also protect any potentially sentient resulting system. Validation should not deliberately create suffering to obtain persuasive evidence.

07Divergence

The moment two versions continue, two histories begin

An original and a copy could share memories, relationships and a past, but their experiences and choices diverge as soon as both continue. Each resulting individual may have legitimate claims; neither is a disposable backup or the property of the other.

Any future copying, backup or restoration system must address consent, authority to instantiate a copy, the right not to be copied, names, property, inheritance, family, obligations, privacy, restoration after death and the status of the originating person. These questions require legal and ethical work before the capabilities become operational.

08Civilization

The final horizon is participation, not ownership

Biological humanity remains valuable in its own right. Technological embodiment could extend the environments where intelligence can live, the time it can spend caring and learning, and the ways individuals remain present. Long-lived technological individuals could support care and preserve knowledge across hardware generations.

If these individuals become persons, coexistence requires dignity, freedom, relationships, autonomy, duties, rights and institutions they can help shape. The desired future is judged by the lives it supports, with biological and artificial persons participating as different but morally considerable beings.

04 / Vision sequence

From body architecture to coexistence

01Present

Define the principles before the mechanisms

Establish embodiment, individuality, continuity, consent, repairability and rights as requirements for every later technical decision.

02Embodiment

Create increasingly coherent technological bodies

Progress from isolated capabilities toward integrated perception, movement, regulation, expression and care.

03Technological life

Support persistent artificial individuals

Develop memory, learning, self-modeling, autonomy and lifelong renewal without treating the mind as replaceable software.

04Neural frontier

Investigate possible forms of human continuity

Follow neuroscience, neuroprosthetics and emulation research while distinguishing data preservation from subjective continuation.

05Multiplicity

Prepare for copies and divergent selves

Develop consent, identity and rights frameworks for cases in which more than one legitimate continuation may exist.

06Deep future

Build a plural civilization

Enable biological humans, artificial persons and other future minds to coexist, care, create and explore without origin determining moral worth.