Dubbed Out: AI Threatens Mexico's Screen Translation Industry
A trade that grew rapidly by localizing streaming content in Spanish now finds producers are doing this themselves with AI
Artificial Originals
AI poses an increasing threat to Mexico’s dubbing industry that currently provides audio translations into Spanish for some 60% of mainstream movies and series, according to a report by Milenio.
The announcement that Amazon Prime will use AI to provide dubbing for 12 series next year has sounded alarm bells for Mexico’s industry, which has grown by 36% in the last six years to reach $66 million. It now supports thousands of jobs ranging from voice actors to directors, technicians, and translators.
“Replacing the workforce with machines would be a major blow to the industry in Mexico and would leave thousands of families without income,” said Carlos Sánchez, general director of dubbing studio Caaliope.
Much like artists all over the world seeking to protect their work from being used in AI training, Mexican voice actors in the industry worry about AI copying voice inflections or intonations developed over their careers.
Dubbing was for years considered a cinematic amenity for children, with those who preferred subtitled representing a more sophisticated viewership that was more willing to read and more interested in hearing the original sounds of actors’ voices.
The tides began to turn a decade ago with the explosion of streaming series, making dubbing an expectation and industry standard for many series, according to this 2014 story - perhaps because greater availability of this content put it on the screens of lower-income people who were less likely to speak other languages.
Will Latin American viewers like AI dubbing? Will they know the difference?
Fake Macri Video Fallout
Argentine lawmakers have proposed legislation to regulate AI videos during political campaigns following the scandal caused by a deepfake of former President Mauricio Macri announcing the withdrawal of his allied candidate for the Buenos Aires legislature. Critics are calling the idea a censorious overreaction.
It’s unlikely that the Macri deepfake affected the outcome of that election - Silvia Lospenatto was soundly defeated, but it has already prompted speculation in the region about how deepfakes could play into critical elections in the future.
Peru’s 2026 presidential vote will likely include enormous amounts of AI-generated content, growing misinformation and exploitation of the digital divide, reports El Comercio.
Similar concerns are rising in Chile, which holds general elections in November. The use of AI in political campaigns is not enough to cause alarm, but remains a potentially enormous threat, Rafael Rubio, an expert in artificial intelligence and elections, told La Tercera.
Mexican man sentenced to 5 years over AI porn
A former college student known as Diego N. had been charged with “human trafficking in the form of child pornography,” for using AI to create nude images of students who studied with him, according to Vanguardia.
President Claudia Sheinbaum in December said his case highlighted the need for specific penal code typifying the use of AI for violence against women.
In December, President Claudia Sheinbaum said his case underscored the need to create a specific penal code provision that classifies the use of AI for violence against women as a criminal offense.
Macro Prompt
Chatter about the coming job-pocalypse
One in four jobs are likely to be affected by AI, says the International Labor Organization, with high-income countries most exposed.
AI is going to wipe out half of white-collar jobs and take unemployment to 20% in five years, Anthropic CEO Dario Amodei tells CNN.
Nah, you won’t lose your job to AI - but rather to a person using AI, says Nvidia CEO Jensen Huang.
My take? I agree with CNN that hyperventilating headlines are page one of the AI industry PR playbook.
AI Drives Sales of Test-taking Books
Per last week’s newsletter, AI cheating at universities has gotten so bad that professors have gone back to the dreaded blue-covered staple-bound collection of thin paper known as the Blue Book, reports the Wall Street Journal (I never thought they were thick enough be described as books).
Whether a quaint thought experiment or an actual effort to bust cheating, I might have to rate the Return of the Blue Book as the century’s most shocking comeback story (with the possible exception of The Karate Kid franchise).
A Sane Take on AI
Tech analyst Benedict Evan’s appearance on The Mad Podcast offers an interesting bird’s-eye view of AI while avoiding both hype and hyperventilation.
He offers some interesting foils that rang true to me, including:
AI today is a lot like the internet in 1995 - yes it’s going to change the world, but no, we don’t know how.
AI deep research is a lot like finding a very smart intern and asking them to write a paper on something they know nothing about - there’s not a single line of it you can trust with your eyes closed.
Agents are all the rage - but do you really trust them to book your flights for you? Do you know what happens when something goes wrong?