From text to video: Sora arrives, the artificial intelligence that generates super-realistic imaginary videos

We have long been accustomed to films made with the help of computers that generate extraordinary visual effects, but we are ready to see productions entirely created byartificial intelligence? A scenario that is anything but remote judging from the first examples shown by Sorathe service that uses AI to create videos.

Using a new model of computer language, OpenAI, the American company of GPT Chathas now announced the arrival of this innovative system for generating movies starting from simple text commands. In fact, to get a video you just need to simply write what you want. It only takes a few seconds and you get it results that are difficult to distinguish from content created in a traditional way.

Technically this procedure is called «text-to-video»from text to video, and evidently requires exceptional computing capacity but this function, instead of being confined to some technological laboratory it is available to everyone. The change of pace offered bygenerative artificial intelligence it is precisely this: access to frontier innovations is within the reach of anyone who wants it.

There are already other similar services such as Runways, Pika, Stable Video Diffusionbut Sora seems to have an edgenot only in the realism of the films but also in the possibilities of narrationfor example by being able to change shots while keeping the context, characters and lights consistent, just as happens in professional productions.

At the moment OpenAI has only announced its new service which is still in a testing phase but the operational launch should take place in the next few months. Currently the main limitation is the fact that the maximum duration of videos is just 60 secondsbut everything suggests that this element will probably be overcome with a version of Sora dedicated to audiovisual professionals.

In fact, it is legitimate to ask what impact artificial intelligence will have on the cinema and television industry if in a far from remote future screenplays, actors, video footage and editing are generated by commands given to the computer. No alarmism, real stars and talented directors will continue to express their abilities, but much of what we see will be synthetic and the closing credits of a Hollywood production will simply remind fewer and fewer professionals.

You will also need to be careful not to fall into the traps of deep fakevideos made with known faces and voices but without the consent of the legitimate characters, a problem that can easily spill over into the most intimate and personal spheres.

In short, it is good that technology evolves, good that it does so quickly and with increasingly wider access, but governing this change is essential and it is also up to us to do our part. It is absolutely premature to hypothesize the end of auteur cinema made by thousands of professionals, but it is better not to brand this scenario as mere fantasy.