RAERAE REACTS · AESPA · EXACT OFFICIAL MV
Rich Man Yellow Claw Remix MV React: The AI Video Spent the Artists to Save the Budget
Yellow Claw rebuild “Rich Man” around restrained tension, heavy drops and a second impact that earns its larger footprint. Then the official video uses Google Flow and Veo to generate performers who are not aespa, mouths that imitate aespa and images whose labor trail is harder to see than the remix credits. A song about carrying your own value arrives in a visual that removed the four artists whose voices create that value. That contradiction is not a footnote. It is the review.

Release file
Released: September 19, 2025
Runtime: 03:25
Artists: aespa — Karina, Giselle, Winter and Ningning × Yellow Claw (remix production)
Exact-video distinction: Official SMTOWN/ScreaM Records remix MV lWIQ72KFgrg, made with Google Flow powered by Veo and distinct from the human-member Rich Man master and other remix audio files
Sound: Yellow Claw electronic remix with slow tension-building, restrained vocal framing, heavy bass drops and a more forceful second-drop escalation
Source trail: SMTOWN official master ·
exact MVDaily+197Total views486.1K
First reaction: the music and video argue
The remix gives aespa’s voices space before impact.
The video removes their bodies and replaces them with generated approximations, turning sonic respect into visual absence.
Yellow Claw rebuild the pressure curve
A calmer opening lets the vocal declaration establish meaning.
The first drop releases tension; the second increases weight instead of copying the same event.
The AI disclosure does not settle the ethics
Naming Flow and Veo tells viewers which branded system produced images.
It does not identify training sources, displaced artists, prompts, human cleanup or who approved each synthetic body.
Karina’s authority survives only in sound
Her attack anchors the declaration while a generated figure receives the visible action.
That split exposes how easily likeness-adjacent spectacle can detach voice from person.
Giselle’s phrasing remains human evidence
Conversational timing carries choices a generated mouth merely imitates.
The video cannot manufacture the authorship already present in the recorded stem.
Winter cuts through the restrained build
Her brightness helps tension remain alive before bass arrives.
Synthetic faces do not explain that musical work; they redirect attention from it.
Ningning gives the drop emotional scale
Warmth makes the later impact feel like expansion rather than machinery turning on.
The image pipeline benefits from that feeling without representing its source performer.
A self-worth anthem deserved visible selfhood
“I am a rich man” rejects dependence and external valuation.
Removing the members produces the opposite visual logic: their voices can be extracted while replaceable bodies perform value.
Generative efficiency transfers cost
The label may reduce production time or budget, but savings do not vanish.
Illustrators, performers, set crews and source artists can carry the lost opportunity while computing and platform companies collect value.
Generated diversity can still be empty
A model can produce many faces and styles quickly.
Representation without authorship, consent or durable employment risks becoming visual inventory rather than participation.
The remix itself deserves separate credit
Yellow Claw’s arrangement decisions are audible and should not be swallowed by the AI controversy.
Criticism can praise the pressure curve while rejecting the chosen visual production.
The ending slogan makes the contradiction louder
“We are enough as we are” claims sufficiency after the video replaced the actual artists onscreen.
The sentence reads less like liberation and more like an ethics alarm the production accidentally installed.
Local metadata must record the method
The attachment should identify this as official generative-AI imagery rather than imply a real photograph or member appearance.
Accurate alt text and description prevent synthetic people from being filed as documentary evidence.
The better alternative was not impossible
Animation, typography, licensed artwork, member footage or a credited visualizer could have served the remix.
Generative replacement was a decision, not the only route available.
Four performers, four kinds of pressure
Karina gives the frame architecture and command without making leadership a permanent monopoly on attention. Giselle can make a line conversational, sly or abrasive enough to puncture immaculate world-building. Winter places a bright, controlled vocal edge through dense production and makes small expressions read at speed. Ningning brings range, warmth and a willingness to let power sound emotionally inhabited rather than merely impressive.
Those tendencies are tools, not cages. Every full-group master routes the same canonical card to all four profiles because participation is factual and should not be decided by thumbnail placement, line-count arguments or whichever member a short-form edit made trend.
What changed every time I replayed it
The first replay separated music from image. The second tracked every moment a human vocal decision was assigned to a synthetic mouth that had not made it.
By the third watch, the final self-sufficiency slogan felt almost accusatory toward its own production method.
What I kept replaying
- The restrained build honoring vocal space.
- The second drop genuinely escalating.
- Karina’s authority detached from her body.
- Generated mouths borrowing Giselle, Winter and Ningning’s labor.
- The closing slogan contradicting the replacement strategy.
The larger industry question
K-pop companies increasingly describe generative video as innovation while avoiding the harder questions: training consent, labor displacement, performer likeness, archival truth and whether audiences were offered a meaningful choice.
Disclosure is the minimum, not absolution. This exact file should remain documented because criticism needs evidence, but future production should hire and credit human visual artists instead of treating replacement as progress.
Where it could hit harder
The generative visuals produce continuity errors and interchangeable bodies that weaken character, rhythm and trust.
The video supplies no meaningful human credit trail for the visual decisions beyond branded tools, leaving authorship deliberately vague.
The generated dancer is not a neutral substitute
A synthetic body can be prompted toward gender, race, age, clothing and movement without any onscreen person negotiating those choices. The apparent performer cannot consent, receive pay, build a career or contest how her body is used later.
That absence is part of the product’s convenience. Calling the figure diverse or expressive does not create the labor relationship that representation usually implies.
Likeness does not require an exact face to matter
The video avoids presenting a direct photoreal clone of one aespa member, yet generated women lip-sync to aespa’s recording inside the official campaign. Viewers are invited to transfer performer presence onto replaceable bodies.
Policy focused only on exact facial duplication misses that broader extraction. Voice, styling cues and branded context can create association without one-to-one copying.
Archival metadata must protect future readers
Years later, a card image may circulate without the announcement that named Flow and Veo. If metadata simply says a woman dances in the official MV, the archive converts synthesis into false documentation.
The local attachment should retain source URL, tool disclosure, synthetic-image description and the fact that no aespa member appears in the generated scene.
The musical drop deserves a human visual answer
Yellow Claw create two stages of pressure with identifiable arrangement choices. A motion designer could translate those stages through type, licensed illustration, hand-built simulation or documented performance.
The availability of strong alternatives makes the replacement harder to excuse. Budget is a production constraint, not a moral eraser.
RaeRae verdict
Yellow Claw’s remix has a smart pressure curve and an effective second drop.
The official AI video undermines the song by extracting aespa’s voices from the selfhood the lyric celebrates.
Bottom line
The AI video spent the artists to save the budget. Yellow Claw’s “Rich Man” remix deserves its musical credit, but the official visual contradicts the anthem by replacing aespa’s bodies, obscuring human craft and calling extraction innovation.