There's a lot to say about cinematic AI; it's worth a post on its own.
This AI-generated music video comes from London-based music producer Pavel Zhigarev. He primarily composes music for commercials at his small studio, Noizlab. He started using generative AI and had the idea to create a fake music video for his favorite band, Limp Bizkit, called “What if Limp Bizkit Were a Country Hit.” The video went viral; the band posted it on social media and it has garnered 1 million views. Fred Durst messaged Pavel and asked him to create the official music video for their track “Turn It Up Bitch.”
“The production of 'Turn It Up Bitch' took 2-3 weeks,” says Zhigarev. “Mainly, I had to learn how to enter the correct prompts and choose the right shots from a huge number of randomly generated images.” Luma.ai was released just as the first cut was being finalized. Zhigarev felt that if he didn't implement it in the video, it would look outdated. AI used included ChatGPT (creating multi-layered prompts), Midjourney and Leonardo AI (image generation), Photoshop AI (replacing/adding elements (removing six-fingered hands and three-legged), Krea.ai (enhancing surrounding elements), FaceSwap (replacing generated faces with artist's faces), Topaz Gigapixel (upscaling to 8K), Runway ML (creating videos from photos), and Luma.ai (creating videos with artist's faces (reducing facial distortion)).
AI is great at genre mashups, but in the hands of a real storyteller, it's even better. “To put the film together,” explains Philadelphia-based filmmaker Tim Simmons, known on YouTube as Theoretically Media and 700 subscribers shy of 100,000, “I took a smorgasbord of the best AI tools on the high seas, and in this story I wanted to show you how to bring them all together into one narrative. I'll share a full rundown on my channel later, but for now I invite you into this story. But I'm warning you: there's a storm on the horizon.” Simmons used Midjourney (imagery), Kling (video generation), Runaway Gen-3 (video), Luma AI Dream Machine (video), Eleven Labs (audio), Suno (music), and Magnific (upscaling).
“The new model is truly groundbreaking. Here, 95% is Gen-3 text-to-video conversion, and the atmosphere and flow are pretty close to what I envisioned,” said @A_B_E_L_A_R_T, the French filmmaker and cinematographer behind “Le Voyager,” who prefers to remain anonymous. He worked with much of the film's content as found footage (like “Blair Witch”). He felt that this kind of fake found footage was a great way to use AI, because found footage is not as controllable as AI-generated images. “In the future, I can imagine two forms of generative AI: one where we have full control and the AI serves us 100%, and the other where the AI works independently, transmitting chaos to open a door to the “other side,” like an exploration of mysteries from an unexplained, independent reality.
On X, I discovered Tokyo-based digital creator Arata Fukae. He created this short movie using a tool that I've featured many times in this column. Fukae posted on X, “I used the latest generation of AI to create something like a music video (although it's messy).” He said, “When you look at people on a large screen, their flaws stand out, but depending on how you use it, you can hide them somehow.”
If you want to learn more about AI and filmmaking, you might also be interested in two other articles published in Forbes earlier this week:
“The Great Catsby” is one of 60 new shows from DreamFlair AI that have been produced using the new Gen-AI tool.
DreamFlare AI Launches Interactive AI-Generated Streaming Service This new service blends technology and creativity to deliver an interactive storytelling experience powered by cinematic AI.
Toonstar uses AI to make hit YouTube series Parker James' YouTube show uses AI to achieve speed, scale, and efficiency not possible across the entertainment industry.
