‘Artificial intelligence used incorrectly will exacerbate mediocrity’
The audiovisual industry debates the pros and cons of the new technology at a forum in Spain, where some of the largest production companies have departments that are developing AI projects
“You will not hear us say that artificial intelligence is a nuclear bomb that will destroy our industry,” said Francisco Asensi, advisor at Spain Audiovisual Hub and moderator of one of the round tables at the 3rd Iberseries & Platinum Industry forum of audiovisual professionals held this week in Madrid, where some participants raised urgent questions about the application of AI to the audiovisual production process.
As the industry’s largest international event on the Iberian peninsula, Iberseries & Platinum Industry used X (formerly Twitter) to clarify that the debate would not focus on creative processes following angry reaction from professionals over the original program, which revolved around the possibility of generative AI to create scripts, increase production value, accelerate schedules and reduce costs. The title and description of the debate were tweaked to remove references to scripts and cost-cutting, although these issues were undeniably in the air.
How to incorporate AI is currently one of the most sensitive issues in the audiovisual industry. It was one of the key issues in the negotiation between studios and screenwriters in the recently concluded strike in the United States, and it has become one of the main issues driving the ongoing actors’ strike. The agreement signed by screenwriters in Spain imposes limits on the use of AI in the creative process so that it does not replace writers and is used only as a tool. But the application of AI to the sector goes beyond screenplays.
According to Álvaro Manzano, general manager of Media Industry at Accenture Spain, it is estimated that 40% of a worker’s total hours could be more productive with the application of AI. It is also estimated that over the next five years, investment in these technologies will increase by 26% in the sector. Manzano did however highlight a number of challenges, especially related to ethics and data security, as these are collaborative models that need to be fed information. “Human intervention will always be necessary,” he said.
In Spain, where the forum took place, some of the largest production companies already have departments that are developing AI projects. Mediapro, for example, is developing projects to generate new business models based on existing knowledge, automation projects to help professionals in different fields, and running a laboratory that works with new technologies, according to Mayte Hidalgo, head of the group’s Innovation and AI Center. Hidalgo added that all of this needs to take into account the necessary ethical framework: “We are at a time when we are building an approach to all this technological capacity. The ones who are going to build it are companies with our kind of commitment.”
Santiago Yuste, Director of Information Technologies at the Secuoya Group, pointed to two practical examples of AI in the production processes. On one hand, it can be used to process and catalog the projects they receive, whether scripts, audio or video. “That way we minimize the time needed to identify projects of interest,” he said. On the other hand, it can be used to suggest locations for shooting a new story: “Our location management system, based on experience, automatically suggests locations to the production team,” Yuste said, adding that this can reduce costs and minimize the time it takes to get a project underway.
Another way that AI can be applied to the industry was explained by Alexandra Falla, director of the Colombian Film Heritage Foundation. Her foundation already applies it to coloring processes, increasing the quality of digitization, subtitling processes and the generation of metadata to catalog their material. But Falla points out that in several of these processes human intervention is still essential, both in colorization which requires the experience of the human eye, and in the review of subtitles to ensure that the original context and language usage are preserved.
Algorithm culture
At odds with the rest of the participants was José Manuel Lorenzo, president of DLO Producciones, who expressed concern about the consequences of AI if there was no appropriate “natural intelligence” behind it, and flagged up the use of predictive systems on platforms to help make suggestions and proposals to viewers. “The fear is that platforms will decide what content to make based on those predictive systems,” he said. “That would curtail creative capacity and that natural intelligence I’m appealing to. Am I going to rely on what a computer tells me about the locations for my series? No, I would have to go and see them with the director and find out if they can work. Hollywood screenwriters and actors have put the brakes on the industry for so long, with one of the contested issues being where artificial intelligence is taking us... You have to look at artificial intelligence carefully. I’m looking at it through sunglasses.”
Regarding the risk of the algorithm culture and the danger of standardizing content, Álvaro Manzano said: “Artificial intelligence, if used incorrectly, will exacerbate mediocrity yet we will think that the result is coherent. It is based on the data we have given it, and there is important training behind it. How we use it will depend on where it takes us.” According to Santiago Yuste, “it’s not that creativity is going to be curtailed, it’s that instead of it going on for nine months, we put it into practice for six.”
Coming at the challenges from a different angle, Lorenzo countered, “While it can help with many production systems, and streamline and improve processes, the problem is when we get into the issue of creativity, which is still human, open, free, and not so directed and predictive. I’m worried that we will be directed by production, that we will be told what to do. Artificial intelligence systems are very predictive, but people don’t know what they want. They know what they want from what they know, but they don’t know what they might want from what they don’t yet know. And that’s where the creator pushes the agenda.”
Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition