Free tools Windows power users keep installed
One-click scans. No signup required.
Some links on this page are affiliate links: if you buy through them we may earn a commission, at no extra cost to you.
The disturbing clip is a July 2024 AI-assisted animation based on Dutch artist Ard Gelinck’s “then and now” celebrity portraits. Gelinck originally made polished Photoshop composites placing stars beside younger versions of themselves; the later video added synthetic movement, turning controlled still images into uncanny embraces.
The claim that “everyone” hated it is headline hyperbole. The clip prompted plenty of mocking and disgusted reactions, but no representative survey established a universal response—and Gelinck himself described the result as frightening while remaining fascinated by it.
What is the video?
It is not a conventional film, studio production or documented celebrity collaboration. The short compilation digitally animates photographs and Photoshop composites so that famous people appear to interact with their younger selves.
The referenced coverage was published on July 20, 2024, following an artist post dated July 17. The clip was described as viral, although the available reporting does not establish a verified view count or the exact date it first appeared.
#1 Best Overall
The apparent hugs never happened. They are synthetic visual constructions made from existing images and AI-assisted motion. No celebrity participation or approval is documented in the cited coverage.
The original artwork was Photoshop-based
Ard Gelinck is a Dutch graphic designer and Photoshop artist known for “then and now” celebrity portraits. His original images combine older and newer photographs, with attention to scale, tone and placement so the two versions appear to occupy the same scene. My Modern Met’s coverage describes this earlier process as photographic compositing rather than an AI-generated one.
Those still images were often sentimental or playful. Earlier examples included Will Smith, Keanu Reeves, Tom Hanks, Jennifer Lopez, Selena Gomez and Amy Winehouse, among others. That list should not be treated as the complete cast of the later video.
Recommended Free Tools
The distinction matters. A still composite can hide imperfect hands, awkward joints or slightly mismatched body proportions because the viewer sees only one frozen moment. Animation forces every face, limb and point of contact to remain coherent from frame to frame.
What AI changed
According to the available reporting, Gelinck “set AI loose” on his existing works. In the referenced post, he said he created 17 of 20 images and that AI completed the remaining portion. That is the artist’s description of the project, not an independently audited production breakdown.
The evidence therefore supports describing the video as AI-assisted or AI-animated, not as a project made entirely by AI. The likely broad workflow involved:
- Gelinck’s existing Photoshop composites and source photographs;
- AI-generated or AI-assisted movement applied to the still images;
- human selection and editing into a short compilation.
The exact model, prompts, animation software, frame-generation method and manual corrections have not been identified in the cited sources. It is not possible to determine the workflow reliably from the finished appearance alone.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Which celebrities appear?
The examples documented in coverage include:
- Emma Watson: shown beside a younger version associated with her Hermione Granger era.
- Will Smith: paired with a younger version linked to his Fresh Prince period.
- Michael Jackson: represented in one reported image through three stages—a later-life version, a peak-career version and a child version associated with the Jackson Five.
Other celebrity pairs may appear in the compilation, but the complete cast and exact number of clips are not independently established by the available material.
Why does it look so creepy?
The discomfort comes from more than a general dislike of AI imagery. The video combines a familiar emotional idea with movement that cannot consistently reproduce human anatomy or identity.
- Unnatural contact: hands, arms and torsos may seem to pass through one another or fail to connect convincingly.
- Identity drift: a face can subtly change age, expression or recognizable features as it moves.
- Unequal realism: the source photographs look highly realistic, while the motion behaves like a synthetic imitation.
- Impossible time collision: adult and child versions occupy the same physical space, creating an unsettling time-travel effect.
- Celebrity familiarity: viewers already know the subjects’ faces and mannerisms, so small errors are easier to detect.
- Sentiment turned awkward: a tender reunion can become invasive, bizarre or even vaguely threatening when the embrace does not obey normal physics.
These are visual and interpretive explanations, not the result of a formal psychological study. The sources document the reaction but do not measure why every individual viewer responded as they did.
Did everyone really hate it?
No. The phrase “everyone hates” describes the tone of the headline, not a verified consensus. The reported reactions include jokes calling the video “the most cursed thing on the internet” and comments describing the premise as awkward and creepy. Those examples demonstrate a wave of negative attention, but they do not quantify the broader audience.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallGelinck’s own response complicates the story. He reportedly called the result “f*cking scary,” while also expressing excitement and fascination about what AI could do. That suggests he recognized the uncanny quality without necessarily rejecting the experiment as a failure.
A more accurate summary is that the video attracted substantial mockery and discomfort online, while the available evidence cannot establish that everybody disliked it.
Is it a deepfake?
It is synthetic or AI-assisted media involving recognizable public figures, but “deepfake” needs qualification. The term often suggests deceptive impersonation—making someone appear to say or do something that really happened. This clip is presented as an artificial artwork or visual experiment, not as evidence of a real celebrity encounter.
“AI-assisted animation of manipulated celebrity photographs” is more precise. The cited sources also do not establish whether the celebrities or rights holders authorized the work, nor do they establish a legal violation. Questions about likeness rights, copyright and publicity rights depend on the jurisdiction, context and specific materials used.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →What the clip actually demonstrates
The project is a compact example of the gap between still-image and motion generation. Gelinck’s original compositions could control the exact pose, lighting and framing. Once movement was added, the system had to invent transitions, preserve two identities, model bodies in contact and maintain believable depth.
That creates a basic trade-off: AI can add motion quickly, but motion exposes errors that a polished still can conceal. The sentimental premise also raises the standard. Viewers are not merely looking at abstract movement; they are being asked to believe in an intimate physical interaction between two recognizable versions of the same person.
The result is not proof that all generative video fails, and it does not reveal the artist’s complete technical process. It does show why a convincing frozen illusion can become visibly artificial as soon as it has to move.
What remains unknown
- The exact AI model and animation software;
- whether the process used image-to-video generation, interpolation or another method;
- the complete celebrity list and final number of clips;
- which frames were manually corrected;
- the project’s verified audience size;
- whether any celebrity or rights holder approved participation.
Those gaps are important because appearance alone cannot establish how a video was made or what permissions were obtained.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
