Games

From storyboard to clip in minutes: how Ai rewrites video production

Through artificial intelligence models it is now possible to produce entire films without a crew (or almost)

by Jader Liberatore

3' min read

Translated by AI
Versione italiana

3' min read

Translated by AI
Versione italiana

 

Last March, OpenAI decided to shut down Sora, the artificial intelligence model capable of generating realistic videos from short descriptions. But the news seems to be a paradox as it comes at a time when the field of AI-generated videos is clearly accelerating with more and more credible and widely distributed content.

Loading...

That of Altman's company, therefore, is a choice that does not depend on technological or qualitative limits but on a combination of pragmatic factors linked above all to the computational cost as well as issues related to copyright and data use. Yet reality tells a different story, namely, that AI-generated videos are not only increasingly widespread but are redefining the way in which content is conceived, created and distributed: the real change, in fact, lies not only in the final products but in the production process.

On social networks and particularly on X, a new trend is emerging, namely that of AI videos accompanied by descriptions in which the authors explain how to make them from scratch. And these are not tutorials but a sort of 'behind the scenes' that goes from the drafting of the subject, to the storyboard and the generation of the final video, through simple prompts. The workflow is standardised: starting with a well-described idea, which, with the help of a model such as ChatGPT Images 2.0, is then translated into a detailed storyboard with sequences, shots and scene descriptions; again with the AI, the appearance of the characters and settings is then generated, starting with the description provided by the user.

In the final stage, however, the images produced are fed to video generation platforms such as Seedance 2.0, which are able to cross textual and visual inputs to create coherent and quality animated sequences: it only takes a few minutes, or just a few hours, to do a job that in the past would have required several professionals and days of work. User @OlivioSarikas on X, for example, posted a video in the style reminiscent of a classic Pixar cartoon, featuring a father and daughter preparing breakfast.

The same author illustrated the process behind the commercial of an imaginary seaside resort called LuxLife Retreats, made with realistic-looking but entirely AI-generated actors.

And then there's @MrLarus, who wanted to demonstrate how AI can be used in the creation of a choreographic video clip: a process that starts with the generation of poses, then processed with Seedance 2.0 to turn them into animated clips.

Content creator @Ciri_ai, also on X, shared the prompts launched on ChatGPT Image 2.0 and Kling AI to create the ultra-realistic video of a live TV broadcast in which a girl appears in the Ferrari paddock during a Formula 1 race. https://x.com/Ciri_ai/status/2053757974825619883

Or @Strength04_X shows how he made the commercial for a serum with an Estée Lauder brand-inspired aesthetic using Nano Banana 2 and Kling 3.0 AI.

However, it is worth reflecting on an important point, namely that all this is not just theory. Trailers of non-existent films or hyper-realistic advertisements created with AI are reality and do not merely imitate traditional audiovisual language but reproduce it at such a speed that they change its rules. But there is an interesting element in all this: the transparency of the process. The authors do not hide the work they have done but explain it by making it replicable: they publish prompts, show mistakes, tell what to do to obtain similar results, creating a kind of collective learning, where skills do not pass through formal paths but through continuous sharing. The risk, however, is to reach a saturation of impeccable but increasingly similar content. The question, therefore, is not whether AI will replace audiovisual operators but what will remain distinctive when production becomes instantaneous: if everything can be generated in a few moments, what makes the difference is no longer the image itself but the intention and the choices of those who drive it.

Copyright reserved ©

Brand connect

Loading...

Newsletter

Notizie e approfondimenti sugli avvenimenti politici, economici e finanziari.

Iscriviti