Mind Unfinished, Self Endures.
NexMe「Aetern」 builds an AI twin from your own writing, voice and notes — one that carries your phrasing, your knowledge and your habits of thought — and keeps it entirely under your control.
It can act as your digital agent: answering for you, holding your knowledge. And under rules you set in advance, it can be handed on as a digital legacy.
Aetern means unfinished — not eternal.
From the Book of Songs, weiyang describes a night that has not yet ended: unspent, unbroken, still continuing. It is not a myth of immortality. It is a more sober promise.
Your thinking, your experience and your way of speaking are consolidated into a digital twin you fully control. It carries your expression and keeps your memory. Permission, access, sealing and deletion are all yours to define.
What twin tools are missing is not a better model.
It is an asset mechanism built around your control. Four things nobody has solved properly.
A lifetime of knowledge scattered across documents, recordings and notes, with no way to call it up
It sits in corners of drives and clouds — unreachable, and never reorganized.
Most twin tools train once and stop — no granular permissions, no legacy handling, no one-click deletion
You hand over the material and lose the leverage.
Fear that your voice and likeness will be cloned and misused
Once it is out there, there is no calling it back.
Existing tools lean toward commercial avatars or resurrection — nothing built around control
The product is the spectacle, not the safeguard.
Five capabilities.
Forge the twin
Import articles, voice, chat history, photos and notes to train an AI twin with your tone and knowledge.
Full control
Pause, edit, freeze or permanently delete at any time. Every authorization is revocable in one click.
Legacy rules
Tiered authorization while you are alive; pre-set posthumous access rules, delivered as a digital legacy.
Memory that accumulates
Keep feeding it new material — it iterates instead of being a one-off static copy.
Safety and compliance
Encrypted biometric storage, deep-synthesis compliance, AI labels on generated content.
The switch is yours — before and after.
While you are alive: you are the only switch
Nothing is published, opened to others or modified without your explicit authorization. Access can be tiered by person and by scope, and revoked at any moment.
After death: delivered by the rules you wrote
You define the trigger, who inherits access, and whether the twin enters read-only sealed mode. No rule is written, nothing happens by default.
Three plans.
Start with a personal knowledge twin, or go further with voice, likeness and legacy arrangements.
Light
- Text and document import
- Private, self-use only
- Export and deletion at any time
Pro
- Everything in Light
- Voice and likeness replication
- Granular external authorization
- Knowledge boundary and forbidden topics
Legacy Custody
- Everything in Pro
- Trigger conditions and witnesses
- Heir permission tiers
- Read-only sealed mode
- Independent custody and handover
Frequently asked.
What is NexMe「Aetern」?
The personal (2C) line of NexMe: an AI twin platform that trains on your own material, stays under your full control, and can be delivered as a digital legacy.
What is the relationship between NexMe and Aetern?
NexMe is the parent brand and provides the twin forging and runtime capability. Aetern is its personal line, focused on control and digital legacy. The full identifier is NexMe「Aetern」.
How do Aetern, Citadel and Stellara differ?
All three sit under NexMe and share the same technology and account system. Aetern forges you, serving you and people close to you, centred on control and legacy. Citadel forges organizational assets, centred on permission governance, audit and on-premise deployment. Stellara forges the commercial front of a creator IP, centred on scaled interaction, co-creation and monetization. In Chinese these three lines are named differently: NexMe 未央 is Aetern, NexMe 知垣 is Citadel, NexMe 星叙 is Stellara - same products, different language.
How is an AI twin different from ChatGPT or another general assistant?
A twin is trained on one person only. It answers with that person’s views and voice, operates within a knowledge boundary you set, and outside that range it does not fabricate or make commitments on your behalf.
How much material do I need?
Text is enough to start. Voice and likeness replication need corresponding audio and video material.
Will my data be used to train models?
No. Biometric data is stored encrypted and original material can be exported in full.
Could my voice and likeness be misused?
Generated content carries an AI label, deep-synthesis rules are followed, and authorization can be revoked in one click.
Can I delete my twin at any time?
Yes — pause, edit, freeze or permanently delete, effective immediately.
What happens after I die?
Whatever you pre-set: a witness trigger or condition-based activation, heir permissions, and optionally a read-only sealed mode.
Can you recreate someone who has already died?
No. We do not do posthumous reconstruction. Restricted memorial use is only available where the person authorized collection while alive.
What you should know before you start.
What you should know
NexMe「Aetern」 does not promise mind uploading and does not manufacture immortality.
The AI twin is a digital agent generated from your material. Its views are not yours.
We do not reconstruct people after death. Memorial use is only available where the person authorized collection while alive, with restricted access, limited conversation depth and an explicit disclaimer up front.
Your voice, likeness and material are not used to train models and are not licensed to third parties.
We follow applicable deep-synthesis and generated-content labelling regulations. Generated content carries an AI label.
Start with what you already have.
You do not need to prepare anything in advance. Tell us what you want to keep, and we will tell you honestly whether the material supports a twin worth building.
Your twin, in your pocket.
Upload material, record your voice, set the trigger rules and heir permissions - all in the app. You get notified first whenever the twin has a new interaction.
How far we have come.
From brand definition to the site going live, then sharing, community, cases and this journey - every public milestone lives here. The full timeline and data placeholders are on the milestones page.
品牌定义完成
母品牌 NexMe 之下确立未央：锻造「你自己」，聚焦控制权与数字遗产。边界：谁才是你的分身由你划定。
官方网站上线
wy.nexme.cloud 上线，中英双语，与母品牌及其他两条线各自独立成站。
App 下载通道开放
统一下载地址开放，面向 iOS 与 Android。
This page shows public milestones; user counts and PV are "TBD" placeholders published only after review. Figures are never fabricated.
What other people built with a twin.
Three scenarios picked from the case wall. The full set, more categories and how to submit your own are all on the case page.
The corpus is not an autobiography, it is how you answer
People stall on what to write about a whole life. We change the task: not an autobiography, but the thirty questions you are asked most often. Those pairs are the first corpus.
- Start from frequent questions, not a complete narrative
- Questions you cannot answer stay blank rather than invented
- Editable at any time; control stays with the person
Status: scenario prototype, not a client endorsement. Outcome figures are added only after permission and verification.
Organising a mother life while she is still here
A typical family-handled case: she is alive but speaking is hard. The method is explicit consent first, then transcription from recordings, confirmed item by item.
- Consent first — handling is not the same as deciding
- Sensitive material is confirmed by her before it is entered
- She can have any passage deleted at any time
Status: scenario prototype, not a client endorsement. Outcome figures are added only after permission and verification.
A monthly memoir in a care home
In care settings the twin often does the repeating: collecting the stories a resident tells often into a booklet, which reduces repetition for carers and makes the resident feel heard.
- Only what the resident is happy to tell
- Nothing about other residents or staff is ever entered
- Family sees the booklet before it is made and can veto
Status: scenario prototype, not a client endorsement. Outcome figures are added only after permission and verification.
These are scenario prototypes, not client endorsements. Authorised, nameable cases are being added.
Turn a one-time visitor into someone you can talk to for years.
Followers drift away; a room does not. We run groups by topic. Free to join, never used for fundraising, no promised returns.
Personal planning
- You are here and want your future self arranged: how to accumulate material, where to set boundaries
- When the twin should step in, and when it must be you
- How this connects to the legacy escrow package, and when to consider it
Family and care
- Handling it for a parent, spouse or elder: which documents, how to obtain authorisation
- Where the boundary of acting on someone behalf sits
- What actually works in care settings, and what went wrong
Legacy and wills
- Custody, trigger conditions, how heirs obtain access
- How this meets wills, notarisation and trusts
- Updates when platform policy or the legal environment shifts
Life education and support
- For practitioners and families: how to start the conversation
- How to accompany someone without causing further harm
- No therapy is provided — experience exchange and referral only
Group codes expire and stop at capacity. When they do, register and we invite in batches.