A familiar face. A familiar voice. Six seconds.
"Long Time No See..."
LTNS is our concept for building a short depiction from your file, a short private video of someone familiar, absent from someone's life into a brief, clearly labelled AI-generated moment. First of all, it can look like made for those who did not have the opportunity to say goodbye. In fact, it's something else...
Initial intention is a calm, gentle experience: the person appears to look towards the user and says "Long Time No See..." in a still and labeled synthetic voice. The words and performance are generated, not an authentic message from that person.
CONCEPT STAGE | ADULTS 18+
The application is not yet open to the general public. Closed 18+ demo on request. This website does not accept video uploads. The website does not collect uploads or health records. A future app will hold an account and a source file only as long as generation needs. That will have its own privacy notice. A soothing experience is a design aim, not a proven psychological benefit or a promise of anything.
WHY WE START WITH VIDEO
A private video can preserve a person's expression, movement and recorded voice together. We want to work from that audiovisual material rather than invent a voice from a photograph. A photo alone contains no recording of someone's voice.
The concept requires suitable, audible speech in the source video. A very short, silent or noisy recording may not provide enough information for a recognisable depiction. We'll try to improve it in the future. Now we focus on digital audio-video files. Feasibility and voice quality still need testing; no exact reproduction is promised. Generation would run on vetted cloud infrastructure.
YEAR ONE: THE SIX-SECOND "LTNS" CLIP
Our planned first offering is a six-second transformation for eligible adults within a supported, controlled process. The clip would contain a brief pause, the softly spoken phrase "Long Time No See...", and a quiet ending. Timing would be refined through research and testing.
We aim for a familiar, composed appearance and an unhurried voice, without dramatic effects. User may nudge pitch, speed, denoise toward memory. They may not invent a kinder person. Any alteration of expression or appearance would be explained. The result would remain visibly labelled as AI-generated for depiction. Not the person. Not their words. Year One does not include any therapy session, of course. User confirms age and ownership, we reject unsafe files. The clip is labelled and streamed privately only. If user is in distress, should stop and use recommended numbers.
We all are humans. The proposed journey: Adults 18+. You confirm you own a solo recording of the person who is gone. We take a still - and only if allowed, a short voice sample. We generate a six-second labelled clip that says “Long Time No See…”. You may adjust a synthetic voice toward how you remember. The clip plays on your account only. It cannot be downloaded or posted as their speech. You can delete it. This is not therapy and not their last words.
There should be no automatic playback, repeated-viewing pressure or obligation to continue. Everything is human-centred. Even pricing. Planned Year One clip should be possibly free in the UK. A later short message may be a small paid step. No prices in the US until that market opens, of course.
YEAR TWO: LONGER, USER-WRITTEN MESSAGES
Our next ambition is a longer video using text written by the viewer: words they would like to hear, spoken by an AI representation of the absent person. The text would be reviewed within the supported process due to human-centred ethics, safety and inclusion.
The message must be identified as the viewer's positive script. It cannot establish what the absent person thought, wanted, consented to or would have said. It must not be presented as their real advice, forgiveness or approval.
YEAR THREE: CONTROLLED LIVE CONVERSATIONS
A later research goal is a bounded real-time conversation with an AI representation, with qualified professional oversight, clear session limits, an immediate stop option and follow-up support.
This would require further research and ethical review before any release. AI responses would be generated by a system, not by the absent person.
A ROADMAP, NOT A RELEASE PROMISE
Years one, two and three describe intended stages from the start of operations. Progress depends on evidence, technical feasibility, funding, permissions and appropriate safeguards. A stage may be changed, delayed or not released.
CONSENT, PRIVACY AND HUMAN SUPPORT
Owning a video does not by itself settle every permission needed to depict a face or voice. The proposed service must assess consent, rights and context, including rejection if there's anyone else in the recording. Being absent does not remove a person's right to privacy or their boundaries. We know that grief can relate sometimes to absent but still living persons. Depicting absent-but-living persons would imply rejection.
Proposed safeguards include private processing, limited retention, deletion controls, no public gallery and no model training on personal recordings without separate, explicit permission. These are design requirements to verify before launch, not claims about an operating app.
Do not upload private recordings to unverified tools or experiment alone. Discuss this with a qualified mental health professional first. Under 18? Start with a trusted adult or youth support service. Choosing no AI at all must always remain a valid option.
A STARTUP WITH AN INTERNATIONAL TEAM
London is our base. The team members are Europeans and Africans, and our collaboration brings together professionals from different cultures and generations.
Prof. Dr Grazyna Czubinska leads social research. Roman Mazur MBA is Director. Inclusive AI means involving different cultural perspectives in the research questions, design, evaluation and interpretation.
RESEARCH AND COLLABORATION
Contact: mental.health@ltns.cloud
Please do not send private videos, voice recordings or health records by email. This mailbox is for project enquiries, not counselling or emergency support.