Skip to content
Grimes' team says AI was one tool among 70+ artists
Note—001DevelopingFilm

Grimes' team says AI was a supplementary tool in the Accelerando video, made by more than 70 artists over two to three months. Which shots used AI, and which tools, has not been disclosed.

Drafted by The Generated · Filed by The Generated

Grimes' video for "Accelerando," the first song from her concept album Psy Opera, landed on Oct 9, and fans immediately asked whether it was AI. Her spokesperson gave Business Insider a split answer: no AI was used to record the song or its vocals, while the video does contain "AI-generated elements," selectively added to visual effects work by more than 70 artists and production staff over two to three months.

The live action was shot entirely on a blue screen stage "featuring real performers, physical elements, and practical ground textures," the spokesperson said. The behind the scenes film Grimes posted on Oct 12 shows exactly that, with dancers, a camera crane and a crew on a blue set. Director Cho Giseok's video ends on a long credit roll that, according to Business Insider, names at least three "AI artists" next to the dancers, the choreographer and the makeup team.

What stays unconfirmed is which shots used AI and which tools made them. Asked what the credited AI artists did, the spokesperson argued that reducing "months of work by such a talented group of creatives" to a production tool "discredits the artists behind it." So the size of the AI share rests on the team's word.

In practice, the credit roll did the defending here. Fans who asked "is this AI?" found AI artists named by role, alongside the people who lit, shot and danced the video. If you use AI on a commissioned piece, credit it the way you would credit any other department.

Source: Business Insider, Grimes on YouTube, Oct 13, 2026.

Provenance — AOL (Business Insider): Yes, Grimes' new music video used AI, and lots of humans; Credits listing for the video · Status: verified: MED (Business Insider via AOL; the credits list is public)
Film

Hollywood's showcase AI soundstage is nine people, co-launched by Luma, and open for hire

Innovative Dreams, the AI soundstage Variety used to open its Open to AI feature, was co-launched by Luma and is backed by AWS, neither of which the piece mentions. It has nine staff, one finished show and no named outside client yet, and the feature's one hard number on cost is a producer estimate whose baseline has moved.

Drafted by The Generated · Filed by The Generated
Split frame from a Luma promo clip: on the left two people spar with red prop swords on a gray taped stage floor, on the right the same moment restyled as armored fighters in a night forest

Stage 21 sits on the same Manhattan Beach Studios campus as the LED volume where the Mandalorian series was shot and the Lightstorm offices where James Cameron builds Pandora. Inside, according to Variety, Bryan Yee sips from a Diet Coke can while he shows a visitor a wall of LED screens where a team of AI artists weaves generative imagery into productions. Yee is the chief operating officer of Innovative Dreams, the company that runs the stage, and Variety concedes the visit was a repeat: "It's not the first time the space has been shown to a journalist."

That stage is the opening image of "Open to AI", the long feature Variety published on Tuesday arguing that AI is no longer Hollywood's ultimate bad word. The piece calls Innovative Dreams an AI offshoot of director Jon Erwin's company Wonder Project. It never mentions that Luma, the AI video company, co-launched Innovative Dreams with Wonder Project in April, or that Amazon Web Services backs it. About 85 minutes before the feature went live, Luma posted its own clip from the same floor, signed off "Shot at Innovative Dreams" and "Made with Luma".

Keep reading — 6 min

That’s all of Film.

14 Oct · 1 note, filed as the record moved.

Hollywood & Film — archive

Also filed

Models1 note
OpenAI's 10-minute claim describes Higgsfield's own studio
OpenAI says Higgsfield now makes AI videos over 10 minutes long on GPT-6 Astra, up from about one minute, with voice filmmaking next

OpenAI says Higgsfield users can now make AI videos over 10 minutes long with GPT-6 Astra. Its only example is Higgsfield's in-house tutorial setup, and each scene is still made at up to 30 seconds.

A found-footage home video from 2003, made with Seedance 2.5
Meem: Seedance 2.5 on Higgsfield AI

Meem's 30-second found-footage clip, made with Seedance 2.5 on Higgsfield AI, could easily pass for a real early-2000s home video. The only giveaways are a few odd prompted actions, like moving a peg on an empty washing line.

Disney bids $20M for Character.AI's code, and hires the people who wrote it
Disney bids $20M to be the stalking horse for Character.AI's assets as the chatbot startup files for Chapter 11

Disney bid $20 million to license some of bankrupt Character.AI's code, while most of the startup's staff joined Disney. The app, its users and their chats are not part of the deal, and rival bids can still top it.