Creator Makes Songs and Music Videos Completely with AI |

In the caption on Instagram, Fukoe lists the tools he used for the creation:

Music: #chatgpt #sunoai
Video: #DreamMachine #Gen3 alpha #sound
Image: #Midjourney #stablediffusion
Edit: Photoshop, AfterEffects

The video features a protagonist who sometimes rides a lion, sometimes flies over the city, sometimes looks out to the horizon with angel wings, and sometimes breathes fire. All of this would have been difficult to shoot with real people, or at the very least extremely complex and expensive. With the help of Gen AI tools, creators can create such videos relatively quickly with practiced guidance. And you can also create the songs yourself, including the lyrics and sounds — even if you know less about music. Tools like Suno AI or Stability AI’s new open-source audio generator make this possible.

From text to sound:

Stability AI launches open source audio generator

© AI generated with Adobe Firefly

There are tons of tools that can help create rich AI content. And brands can use it for advertising purposes too. Toys“R“Us recently launched its first campaign created with OpenAI’s Sora toolThe . Video is intended to tell the brand’s founding story. It is the brand’s first brand film.

The possibilities of AI creation seem hardly limited and also allow individual creators, smaller companies etc. to one day create elaborate videos, songs and the like for their own purposes. They require much less resources. However, there are also risks associated with this kind of content creation.

Issues that may arise while creating AI

With the big AI video tools like Sora, Gen-3 Alpha and Co. it is not clear where the training data actually comes from. With Sora there were already indications that OpenAI was easy public YouTube video creators used as a base could have. Anyone who develops their entire creative work with Gen AI tools runs the risk of, at least implicitly, seizing the intellectual property of third parties in the broadest sense of the word. This should also be the case when creating song lyrics or sounds. Legally, this creation is still in a gray area. However, artists often complain about training AI tools with publicly available content, from books to songs to images, because their rights as authors and creators could be violated. This often cannot be understood specifically in the context of AI generation.

Another problem in the context of the fact that content is so easy to create can be that social media and other platforms can be flooded with similar content. In the text part, for example, Google has already done that with new spam guidelines responds to widely generated AI texts without quality standards; nevertheless, some AI copy is currently ranked well. Although there are AI labels on social media, which are intended to provide information about the creation of the contentwill likely see a lot of content enter the platforms under the radar, leaving users with the task of differentiation.

Ultimately, AI content could take away reach and revenue from creators, as ads can be placed even in the context of well-performing AI content. Whether advertisers will be successful in this is questionable. In this context, there could also be a brand suitability issue if the display occurs in this context.

The many risks associated with creating Gen AI content are a side effect of the rapid technological development. They do not diminish the potential that this development brings, but creators and brands should always be aware of them. As in many contexts, transparency can be crucial in this regard. Arata Fukoe also clearly shows how and with what his music video was created. But it is also clear that we must prepare for a time when AI content will increasingly permeate the image of digital media.

Dream Machine AI videos:

Sora Contest available to everyone

© Luma AI, screenshot from the Luma AI Dream Machine website.
© LumaAI

Source link

Leave a Comment