An artificial intelligence creates films according to the mood of the audience

Can you imagine if it were possible to stand in front of a film and interact with it, making it change course in real time and according to the emotions the viewer feels while watching it? Well, this is no longer a possibility: “The artificial intelligence (AI) we have developed achieves this. If the system detects that you are not laughing in a film that is supposed to be funny, it will analyze your social networks and other data to try to make you laugh and the film will change its narrative direction automatically”, says Carlos Fernández of Vigo, the director of Emotional Films – which has developed this AI – and director general of the Navarre studios of animated films and video games Dr. Platypus

We live in an era in which the arrival of AI is already a reality and, almost without realizing it, it is entering our lives: “A very strong technological disruption has arrived and the way we live and relate -us with the world will change little by little with the arrival of AI, not only in the cultural industry, but wherever there is human activity”, Fernández de Vigo points out. For this reason, the director of Emotional Films believes that it is necessary to lose the fear of experimenting with this new technology: “I have been working with AI for five years and in R&D projects for twenty years, and I can assure you that I have rarely seen such clear the deep disruption that this technology presents. People are not aware of everything that is coming. There is talk of ChatGPT and image generators, but all this falls short, it is only the tip of the iceberg”.

It is in this technological disruption that the project director has seen a streak of market and innovation in which, after five years of work and with the collaboration of more than 40 researchers, films have been developed that adapt in real time to the viewer’s emotions: “Through a complex system composed of ten interconnected AI models we manage to generate and adapt the film. The adaptation of which our AI is capable reaches the extreme that the content is never repeated, even if the same viewer watches the same film a hundred thousand times”, affirms Fernández de Vigo.

Mikel Galar, technical director and head of Emotional Films at the Public University of Navarra (UPNA), explains that these AI models, in order to be able to detect emotions, are trained with “images of people, of faces that express different emotions and thereby produces a database free of bias. In this way, the AI ??model is effective with anyone who uses it.”

Galar emphasizes that these databases are usually irregular: “There is usually more representation of certain races, ages, genders, of certain stereotypes… This explains why there are people with whom AI does not work well, because the basis of data is not representative of the features of the viewer”. For this reason, it says that, to train its AI models, they have “dedicated a lot of time and effort to creating a database that is as free from bias and stereotypes as possible”.

In addition, Galar points out that in order to fully enjoy the experience of Emotional Films, the viewer must collaborate with the technology: “For this to be possible we must have access to different parameters, the most impressive is the of the viewer’s emotions, but we use others: the location from where the movie is being watched, the age or gender expression of the person in front of the screen , the social context of the viewer and everything that surrounds him, the day and time of year, the weather, and even if it’s day or night…”.

Galar assures that access to all these parameters makes possible a transparent interaction with the film of which the viewer will be little aware, but which will deeply influence the artistic aspects of the experience. In this sense, and as Fernández de Vigo states, it is not the same to see a romantic film on February 1st as to see it on February 14th, and it is also not the same to watch the film alone or with company . The AI ??will take all these details into account and use them to instantly tweak it and create as personalized an experience as possible.

On the other hand, it must be taken into account that the films are personalized within a context and decisions previously established by the screenwriters. The latter will not have to write the dialogues of the scripts, but – as the project director explains – will have to limit themselves to defining the characteristics of the characters and the AI ??will do the rest: “Their job will be to create the character itself ; create it with concrete experiences, with a particular cultural and emotional context, with its own temperament… From here the character will develop the dialogues and actions throughout the film, taking into account the different parameters of the spectator”.

Screenwriters, like directors and producers, “must adapt to the changes that AI is introducing in their discipline”, says Carlos Fernández de Vigo. Although the director, who is a self-proclaimed “film fanatic”, claims that the technology of Emotional Films is not intended to replace it, but rather “is another way of approaching artistic creation , a way to generate a more intimate connection with the viewer”.

Exit mobile version