Matthieu Lorrain’s Post

View profile for Matthieu Lorrain, graphic

AI & Creativity, Google DeepMind

🎶🎸This music video was entirely created with 7 different AI tools... 🚀 GenAI is increasingly allowing for better control in the hands of talented creatives thanks to new advanced workflows... 🛠 In this case, this video created by Arata Fukoe, was made with the following tools: - Music : ChatGPT, Sunoai - Video : DreamMachine, Runway's Gen-3, Kling - Image :MidJourney,Stable Diffusion - Edit : Photoshop, After Effects 👨💻 I often hear some commenters reproaching the lack of control and human craft in genAI creativity. Yes you can quickly make short sequence with a single text prompt and little effort. And yes we are still very early in the development of those tools... 👉 But the reality of professional AI production is that it requires today complex pipelines making the most of each tools to increase your ability to control the image, its ingredients, the lighting, etc.. 🎥 Most "AI production" combine AI tools with traditional technics (from live action footage to "traditional VFx"). This is what i call "hybrid production" 🎞 💻 In addition, "technology is not an idea" and genAI shall remains a tool. It should not replace most of human creativity, great stories, amazing concepts, beautiful visual curation or overall skills...all the things critical to stand out in an age of abundant content creation... 🙏 I hope my fellow creative professionals could avoid thinking in terms of binary "AI or not AI" and dive into more nuanced discussions. The integration of genAI into our creative development is more of a imperfect, hybrid process, combining pre-existing tools and methods with new features... 👀 What do you think? How can we make the most of those new AI tools while preserving the craft and skills our creative community has been honing for decades? Creativity is liquid 💦 #genai #aicreativity

Ben Feuer

Ph.D Candidate | NYU | Deep Learning

1mo

Photoshop and After Effects are not AI tools, they are complex software packages used to create and manipulate visual content. This video could have been created with Photoshop and AE ten years ago.

Stefan Kirschnick

Sr. Insights Analyst LinkedIn | GenAI Explorer

1mo

an AI video not trying to replicate reality but embracing the possibility to create "weirdness" and the "new" :) love it. finally. Super well done.

Roger Belveal

UX Architect, Futurist Artist, Inventor, & UX Activist

1mo

Meaning matters more than it did before. In a world where anything is possible visually, the what matters less and the why more.

Personally, I think it's a mess. It's very disjointed like most other AI works that I have seen. A whole lot of checking boxes to show what it can do with no cohesion. Last time I checked, audio/lip sync was sort of important for music videos. Not so much here...

Richard Buchanan

Strategist | Facilitator | Therapist @ Fellow Progress

1mo

I love it, yet sometimes I feel like AI video is like looking in a kaleidoscope. Do you know what I mean? When I was a kid I would pick one up and be enamoured, yet after a while it felt like I was overwhelmed by the complexity of it all.

Robert Sawulak

Emergency Services Graphics & Branding Production at Endura Ltd

1mo

"I hope my fellow creative professionals could avoid thinking in terms of binary "AI or not AI"" As soon as genAI stops being trained on the work of artists scrapped without their consent or recompense, you might get close to that conversation starting.

Damon Lenahan

Frequency and AI Technologies

1mo

No soul. . Canned lyrics. Eye candied senseless visuals. Hard linear melody that plodds along. As they say, "you are what you eat". We will become as empty as the products we create with this stuff

Lev Manovich

Theorist of digital and AI cultures / Author of 17 books on AI aesthetics, digital art, media theory, digital humanities and cultural analytics / Presidential Professor, Graduate Center, City University of New York

1mo

Beautiful video and excellent post. Yes, hybrid is the future. And lets remember that hybrid combinations is in fact how was already made for for many decades. Optical printing in 1940s. Physical models and computer camera control in 1970s (Star Wars). CGI and other special effects in 2000s.. so GenAI is just another technique added to this long list. Cinema has always been hybrid!

Michael Simo

Content Production, Creative Strategy, Founder, Auteur Theory

1mo

Agreed. It’s a collaborative process. I am a director, videographer and editor, and I do VFX/Playback on Nonunion commercials. Working a lot with the VFX artists, working on sets that use compositing and special effects to and heavy post, I see the utilization of AI as a tool, the way we already seeing it being rolled out in Adobe Premiere and other Adobe cloud services. The ability to remix a song just by dragging it to the desired length (a several hour process condensed into 3 seconds), auto-cut an entire clip of mixed footage with almost 100% accuracy, and to circle an object in frame and have it automatically remove it by keying it out of every shot and replacing it with AI-generated images/textures is wild already. I see editors having a much bigger role in the process. But there are also union rules and certain roles are going to be separated. I wa talking about with with my buddy in Local52 he said there’s already negotiations happening and at the end of the day, whos controlling the image we see on set? It lies somewhere between the DIT and the VTR. If Qtake integrates with some AI tools or Adobe gets into live shooting software, and they integrate AI tools to generate composite backgrounds on the spot, etc.

Like
Reply
Sharon Barnes

Executive Producer - Leadership, Story, Communications, Strategy (former White House)

1mo

I like to think about collage. The quality of the pieces are important, but how the whole is put together is most significant. It takes craft, vision, and meaning to make something unique and connected. Sometimes you want something rough in one moment and something crystal clear in another. The way the music interplays with the visuals, the pacing- all of it- that's intuitive. There are 3 images in this work that stand out to me. I am interested in the development of the technology.

See more comments

To view or add a comment, sign in

Explore topics