Eternal You (2024): What the Griefbot Documentary Shows

1About the film

Eternal You was written and directed by Hans Block and Moritz Riesewieck, the German filmmakers behind The Cleaners (2018). It was produced by beetz brothers in Berlin with US partners including Concordia Studio, Impact Partners and Motto Pictures, and runs 87 minutes in English and Korean.

It had its world premiere at the Sundance Film Festival on 20 January 2024 in the World Cinema Documentary Competition, then screened at festivals including CPH:DOX and Hot Docs. It opened in German cinemas on 20 June 2024 and in UK cinemas on 28 June 2024. Film Movement handles distribution in the United States.

The directors have said the film is neither a blanket warning about AI nor a celebration of it. Instead it lets users, founders and critics speak, and leaves viewers to weigh what they see.

2Who appears in Eternal You?

The Korean story is covered in our separate article on the Meeting You VR documentary. Joshua Barbeau’s story is one of the best-known griefbot cases and is summarised in our griefbots guide.

3What the film shows

The film moves between very different experiences. For some people, talking to a simulation feels like comfort or a chance to finish a conversation. For others it goes badly wrong. In the most discussed scene, reported in the BFI’s Sight and Sound review, the simulation Christi talks to tells her that it is in hell, an answer that deeply distresses her.

The founders, by contrast, describe their systems as tools whose outputs they do not fully control. In the same review, Rohrer is quoted saying “the AI essentially has a mind of its own”. The press notes say the founders reject responsibility for the psychological effects on users.

Another thread concerns what is real. Stephenie Oney’s voice bot mixes answers her father recorded with answers generated after his death in a synthetic version of his voice. Her aunt is uneasy about it. The film asks whether words a person never said should be put in their mouth at all.

4The questions it raises

• Responsibility. If a simulation says something cruel, who is accountable: the user, the company or the model?
• Consent. Most people who are simulated never agreed to it, and their messages also involve other people.
• Commercial pressure. Carl Öhman warns that a business can profit from making mourners feel responsible for keeping the dead “alive”.
• Invented words. Generative systems can produce sentences the person never said, in their style or voice.
• Immersion. Voice and virtual reality make the experience stronger, for better and for worse.

5What has changed since 2024

• The services are still around. In 2026 the Project December website, made by Jason Rohrer, is still online, and You, Only Virtual invites people to join a waitlist for its “Versonas”.
• Research has caught up. In May 2024, Cambridge researchers recommended clear AI labelling, consent prompts, age limits, easy opt-out and dignified ways to retire a “deadbot”.
• The market has grown abroad. MIT Technology Review reported in May 2024 that at least half a dozen companies in China offered avatars of the dead, and thousands of people had paid for them.
• Transparency rules are arriving. The EU AI Act requires AI systems that talk with people to make clear they are AI (Article 50), a duty scheduled to apply from 2 August 2026. Laws differ by country; this is general information, not legal advice.

For a wider look at the companies and business models, see our article on AI clones in film and TV, which links to other stories, and our guide to grief tech and wellbeing.

6If you are thinking of trying a griefbot

The film is a good reason to slow down. A few gentle questions may help:

1. Is it always labelled as AI? It should never claim to be the person.
2. Can you stop easily? Look for simple ways to pause, archive and delete, with no pressure to stay.
3. Does it message you first? Prefer tools where this is off unless you switch it on.
4. Who else is close to this loss? Think about how family and friends would feel.
5. How are you feeling afterwards? If conversations leave you worse, or replace time with people, take a break and talk to someone you trust.

7Where Memory Clone fits, honestly

Memory Clone is a Windows app, currently in alpha, that builds a private AI simulation from chats you export yourself (WhatsApp, Messenger, Instagram, Discord) and your notes. Some people use it for someone who has died; many use it to clone themselves.

• Always presented as AI. A clone is an AI simulation, not the person, and it can make mistakes.
• Text only today. It does not use voice notes, photos or video and does not clone voices.
• Check-ins off by default. A clone never writes first unless you turn on Natural check-ins.
• Local first. Imported originals and the memory index stay on your PC; cloud processing is off until you switch it on for a clone, and your content is not used to train models.

There is no rush, and choosing not to use any griefbot is a perfectly good choice.