A human model, built with permission
Most AI learns information.
Lumina Twin learns a person.
Lumina Twin turns deliberately approved memories, reflections, activity, documents, and voice into a private, evolving Twin—while the person remains in control of every source and every guest.
The difference
A chatbot recalls facts. A useful Twin must also model values, preferences, communication style, relationships, and uncertainty.
Grounded in approved evidence Transparent about uncertainty Revocable by designJudge walkthrough
Evaluate the real product, not a concept video.
Follow these steps in order. Each one identifies the capability and the evidence visible in the live system.
Consent before collection
Open Collectors and Connected Sources. Every source begins off and explains its permissions before activation.
Open product →Deliberate human signals
Answer a Daily Question, save a memory, or approve a document. The owner selects privacy and can review the resulting record.
Try Daily Questions →An evolving Twin Brain
Open Twin Brain to see confidence-ranked values, personality, preferences, worldview, and interests derived from approved evidence.
Inspect Twin Brain →Human conversation
Ask a personal question, then a follow-up. The Twin carries recent context, matches emotional tone, avoids database language, and stays within its evidence.
Test Twin Chat →Voice with explicit consent
Voice enrollment is separate and optional. The owner approves samples before synthetic voice playback becomes available.
Review Voice controls →Controlled family access
Create a trusted guest key, choose the relationship and allowed topics, then revoke it. Guests never receive dashboard or raw-data access.
Open Guest Chat →Exit is a product feature
In Privacy controls, review consent history, export data, delete individual records or categories, revoke tokens, or delete the account.
Read the privacy promise →Safety architecture
Permission is not a checkbox. It is the system boundary.
User chooses a source→Permission is recorded→Evidence is reviewed→Twin learns→Owner may export, revoke, or delete
Lumina Twin is an AI simulation and memory interface—not consciousness, legal authority, or a replacement for a person. It is not intended for emergency, medical, legal, financial, or other high-impact decisions.
What if the most important human knowledge is not found on the internet?