Home / Kira World / Person Creator

Synthetic individuals / character source rules

Person CreatorExperts, companions and characters.

Person Creator is Kira World’s chosen public name for its synthetic-individual creation research. The tool prepares identity, source permissions, interests and characteristics for different kinds of people. Temporary or continuing describes each individual, not the creator itself. A verified one-click complete synthetic person is not yet available.

Warm futuristic virtual apartment with person and holographic digital companion; illustrative Creator and Kira World direction
Illustrative scene of a human and holographic companion; not an operational automatic character creator.

Kira Labs decision · October 24, 2026

Temporary Creator is now Person Creator.

We adopted the new name because not every individual this system is intended to create is temporary. The early project name described only one possible use. Its broader aim is to prepare distinct experts, companions and characters, some for short-term tasks and others intended to develop continuing identities, experiences and relationships.

The name and public descriptions have changed, not the results: mind/body/voice integration, meaningful continuity after restart and live unscripted interaction still need qualified testing. Historical research titles, image assets and internal source identifiers may retain the former name to keep old evidence and links intact.

Evidence by stage

Current work and clear limitations.

These are separate engineering stages, not a single overall completion score.

Existing foundations

Identity/profile plans

Name, background, interests, source history and separate memory plans can be prepared for controlled experiments.

Different source types

Fiction, history and original experts

Fan-made, historical and original research profiles have distinct provenance, permissions and test claims.

Unfinished

Automatic mind + body + voice

Full self-directed learning, natural voice and complete individualized anatomy have not been accepted.

Why this area matters

One creator, several source rules

A historical simulation must cite sources; a fan experiment is not an official licensed character; an expert prototype does not hold professional qualifications.

  • Keep histories and privacy permissions individual.
  • Label fan, history and original experts distinctly.
  • Test actual running characters before release claims.

A source review is not a verified working feature; references and illustrations remain labeled.

Planned Person Creator source, body and voice flow infographic
Original October 22 planning illustration; its artwork still uses the historical Temporary AI Creator name. Not all stages are running.

Original detailed project research

Full records and technical evidence

These sections were moved intact from the former long Kira World homepage. Findings, photographs, safeguards and unfinished tests retain their original descriptions and source boundaries.

Person Creator · How a new synthetic person is prepared

Request a person.
Begin their mind, body and voice.

When someone requests a new AI individual, the intended Creator workflow starts the person's mind and identity profile, then sends the body description to Avatar Builder and the voice characteristics to the current voice-generation system. Those body and voice paths can be prepared alongside the mind, before any combined character is accepted in a tested runtime.

Illustrated proposed creation workflow: a person request; a developing identity and mind; body details sent to Avatar Builder; voice details sent to a local voice generator; integration and testing; a possible synthetic person in Kira World. An upper panel depicts an unbuilt future virtual voice box and vocal cords after NewBrain research.
Illustrated system design—not a demonstration of a finished creator. The mind, body and voice branches represent the intended integration. The virtual vocal cords and voice box are future research, not current runtime components. View the detailed graphic ↗
01 / PERSON REQUEST

Who is being created?

Name, identity, background, interests, personality, visual traits and desired voice. Sources and permissions remain attached to the request.

02 / MIND PREPARATION

Initialize an individual mind

Current creator foundations prepare the identity, source pack, behavioral profile and memory plan. A verified reusable NewBrain learning backend is a future integration.

03 / TWO PARALLEL HANDOFFS

Body → Avatar Builder, Voice → Generator

The design routes appearance, proportions and expression to Avatar Builder, while pitch, timbre, accent and reference permissions go to the existing voice generation pipeline. Both handoffs need full end-to-end integration tests.

04 / FUTURE PHYSICAL VOICE

Virtual voice box and vocal cords

Later research aims to simulate airflow, vocal-fold vibration and articulation for the person's natural-sounding speech, coordinated with facial motion. This is planned after suitable NewBrain and voice-system qualification; it is not working today.

Current development boundary: candidate creation and voice-profile infrastructure exist, but this page does not claim fully automatic mind training, personalized 3D anatomy, a one-click working body/voice handoff, or synthetic vocal cords. Successful integrated conversations, body movement, mouth synchronization, persistence and permissions still require recorded testing. The diagram illustrates the direction and reuses a conceptual character image, not a runtime screenshot.

Person Creator · Historical development examples

From an expert profile
toward a continuing individual.

Robert used the original Temporary AI Creator identity and expert-profile workflow for original synthetic individuals, fictional experiments and historically bounded character studies. Each project requires its own sources, permissions and later testing.

Featured example · In development · Planned free public release

Evelyn Hart AI and technology history specialist

Evelyn began with the earlier Temporary AI Creator workflow (now Person Creator), which Robert used to develop her original expert identity, background, interests and knowledge focus. She is now being developed further as a continuing synthetic individual: someone intended to read, hold conversations, develop interests beyond her specialty, and retain permitted memories of people and experiences.

The plan is to make Evelyn available to the public for free once her standalone application is ready and passes real conversation, authorized voice, visitor-memory, installation and body-integration checks. She is not yet downloadable and no release date is promised.

What has actually passed: bounded head/eye, software and library tests, plus saved reference-body renders with an outstanding side-framing failure. A fully working new mind, natural speech, integrated body motion and secure long-term conversation memory still need end-to-end verification.

Promotional concept portrait of Evelyn Hart holding a tablet beside holographic research panels. This is an illustration, not a captured running application.
Concept art inspired by Evelyn's modeled head · Not a runtime demonstration

Other characters and expert prototypes

Different people, different source rules.

These examples demonstrate character/profile research and controlled testing. They do not mean every profile has a finished standalone mind, 3D body or generated voice.

Fictional character study

Peter Parker

A fan-made, unofficial fictional-character experiment exploring personality, fictional continuity and character-specific knowledge. It is not affiliated with or endorsed by the character's rights holders, nor a public licensed character release.

Historical research

H. H. Holmes

A source-limited simulation set around late 1893. An archived, deterministic 12-scenario guard test checked chronology and unsupported action claims; it did not establish a natural conversation with the real historical person.

Original expert prototype

Emily Carter

A programming-focused candidate with an original identity, background, topic plan and personality interests. Professional expertise, teaching outcomes and real-time factual reliability are not yet established.

Original expert prototype

Jessica Hale

A robotics-focused candidate used to explore specialty knowledge, individual interests and relevant source packs. The profile is a development example, not a credentialed roboticist or qualified autonomous research agent.

Possible extension · Digital Twin research

The same Creator could help build an authorized digital twin.

With someone's explicit permission, the creation workflow could be adapted to start from their real-world source history rather than an invented biography. The goal would be a clearly identified synthetic representation grounded in approved records—not a claim to be the original human being or to possess the human's lived experiences.

  1. 01 / Approved personal materialsUse consented interviews, writing, images, videos, biographical records and memories, with privacy and source-attribution rules.
  2. 02 / Mind, body and voice profilesPrepare a distinct identity/memory model; route permitted appearance details toward Avatar Builder and authorized reference audio to the existing voice-generation system.
  3. 03 / Controlled continuityTest accuracy, identity boundaries, consent, corrections and later memories across restarts. Embodied movement and a virtual vocal-tract voice box remain future research.

Existing related work: Kira Labs' Synthetic Robert research already studies owner-authorized memory context and an approved self-voice route. Turning the full Person Creator into a one-step digital-twin builder is not implemented or verified. It would require additional consent controls, safe export rules, qualified voice and body integration, and realistic testing.

Read the dated expert and historical-character evidence → · Explore Synthetic Robert's existing limits →

Read the original research

Records and development history.

Detailed results remain available on their original Kira Labs pages. Private source media, user histories and uncensored anatomical images are not published here.