“The day has come: they replaced me in a job with a voice generated by artificial intelligence,” Alejandro Graue, a 36-year-old Argentinean announcer and dubbing actor, wrote on Twitter. He had been dubbing a famous YouTube channel into Spanish for a couple of months with millions of subscribers in English, of which he prefers not to give the name: “The author is a histrionic boy, with some very installed idioms, sometimes stutters, who stops in middle of a sentence and start another. He is something that requires acting technique, to be able to represent him as well as possible, ”he explains by video call to EL PAÍS.
In January he saw that the channel continued to be updated with videos that he had not dubbed: “When I hit the play a voice totally lacking in sentiment was heard saying: ‘Hello, welcome back to this program, today we will be…’. Something horrible, ”she recalls. In the comments people complained about the change, but for now the videos have continued to appear with that artificial voice, which is incapable of reproducing the particular way of speaking of the youtuber.
Then he posted the tweet, which has already been seen by more than 740,000 people. The problem for Graue is not that job that he has lost. He is a freelancer and has others, at the moment. The problem is the trend: “It is worrisome. We are on the way for the human factor to disappear from something artistic, ”he says.
Yyyyyy the day came: I was replaced in a job with a voice generated by Artificial Intelligence. Thanks to all the actors and actresses who give their voices to create this shit that will eventually make us all obsolete. The artificially intelligent ones are you.
– Ale Graue (@Alegraue) January 10, 2023
In just a few months the voiceover industry has been disrupted by artificial intelligence. The explosion of ChatGPT and applications to create photos or illustrations since last summer has also reached the voice. As in other creative trades that are also threatened, there is still room to reach the moment where the machine can do everything. But no human light is seen at the end of that road: “For two or three months our billing has dropped,” says Noemí Gutiérrez, director of the dubbing company Voces en la Red. “The announcers call us, write to us because they have noticed the drop. Platforms that provide synthetic voices for free have multiplied. There are people who are worth a cheap voice and they throw away with that, ”she adds.
Although there are still those who have not fully noticed the consequences, they can be intuited, according to Begonya Ferrer, a dubbing actress: “As I work for many people from many countries, I cannot know if they have replaced me with an AI in some company,” she says. . Yes, they have cut her participation in a specific type of videos, which now “use artificial voices to make a first test, or voice draft in off, and then they pass it on to the announcer, who doubles over with his voice. Thus, perhaps, they avoid paying for two locutions ”, he explains.
Quality and elaborate intonations still pay off, but many in the industry warn that the machines will continue to hone their capabilities. “I understand that it is part of the technological evolution and that there are a lot of things that can be solved with artificial intelligence, from answering machines to messages that one hears on the subway, in which feeling is not necessary. But everything that is acting, it seems to me that it should not be done; for a matter, even, of preservation of the trade”, says Graue.
Only programmers will be necessary
The danger that the actor sees is that, in more and more trades, humanity is left in the hands only of those who talk to machines: “The person who performs the task will not be necessary. The only people needed will be programmers and that’s it,” he adds. In the debates that Graue has had since his viral tweet, some have told him to forget about “preserving the job”, that this is like electricity: “There are people who compared it to the appearance of electricity and the work of the lamplighters. I said no, that the appearance of electricity was for a global benefit. What about synthetic voices, how does it benefit the public? It only benefits the company that pays,” says Graue.
Nor is Alejandro Graue old enough to think that he could retire soon and forget everything: “If I were 70 years old, I would say that I can’t worry so much about the future because I don’t have it. But it will affect me directly. I can open up to other fields, but without stopping thinking: ‘I have to dedicate myself to something in which I know that they cannot replace me’. Because I have spoken with colleagues, journalists and screenwriters, and everything is beginning to be replaced with artificial intelligence ”.
That rapid spread between trades causes uncertainty. It is difficult to observe the present in peace while machines begin to do, in seconds, tasks that required special skill and years of study or learning.
“When it happened to me, I had been watching it on social networks, with the illustrations and graphic design,” says Graue. “Then we started to look at it much more closely. Of course there is fear. There are those who say that this is only going to replace those who do their job poorly, and that those who have quality will continue working. But there are also those who are directly looking to dedicate themselves to something else. There is everything. I personally believe that, little by little, it will gain ground ”, he adds.
Graue is confident that the help, if it arrives, will not be from a company, but from the law: “I hope there will start to be some kind of regulation. I understand that in Argentina at least work has begun on a project to regulate it; or at least, to require certain sectors not to replace the voices of announcers or actors with machines”.
Brad Pitt speaking with an Andalusian accent
Some companies in the dubbing and voice industry, meanwhile, are looking for ways to curb the loss of projects with technologies that preserve a part of the voice actor’s work. With today’s tools it is relatively easy to clone or create a timbre, which is the specific sound of the vocal cords. But it is more complex to specify rapidly changing tones and intonations, to express surprise, excitement and anger in just two sentences. “Right now you can get a voiceover with Brad Pitt speaking with a Welsh or Andalusian accent, as long as you introduce a prosody (rhythm, way of speaking) into the system with those accents, interpreted by an actor”, says Javier de Alfonso, founder of Voices on the Net.
”We are buying time. Now six months seems like an infinite time to me”, says de Alfonso, while taking a course to adapt to this new version of his job, which is no longer the same: “We are learning and making great efforts to orient ourselves to agencies with a pace of very demanding recording. We have tested several platforms, including a very advanced one that has more buttons than a 747 ″, he adds. It’s hard to shake the feeling that this reorientation is like leaving a handful of floats on the beach while you hear the tsunami in the distance.
This small momentary lifeline now allows dubbing an entire film with just two professionals, keeping the actors’ original timbre: “It’s already something disruptive,” says Noemí Gutiérrez, director of Voces en la Red. “The system modifies the voice of the doubler, giving it the same timbre as the original actors. Thus, a single voice actor can do all the men; and a woman, all the women and also the voices of children”.
“What has not yet been achieved with synthetic voices is that they do emotions well,” says Gutiérrez. “It will still take time,” he adds. In a future in which this technology is fully developed, there would still be a certain space for dubbing actors: some will be able to sell their doorbell; and others, with a less pretty timbre, will be able to sell their acting ability for artificial intelligence to copy.
Subscribe to continue reading
Read without limits