AI has gained a lot of traction in the media & entertainment Industry with a new application seeming to appear every day.
For instance, AI-assisted compression and statmuxing are helping broadcasters make more efficient use of their spectrum. It is also identifying faces, objects and the written word in video frames to enrich metadata with payoffs ranging from more relevant news creation to increased revenue from content licensing fees.
AI is also making it easier to build replay lists for sports productions, create interpolated video frames for synthetic slow motion; auto-mix audio; unearth relevant content and make video shorts based on what’s found for social media distribution; make content recommendations to streaming service subscribers based on their preferences and thereby reduce churn; match ad inventory to content on a contextual basis; and optimize content delivery network traffic (CDN). The list goes on and on.
One GenAI application in particular recently caught my attention. Sinclair, working with generative AI language editing company Deeptune, ran a trial of real-time English-to-Spanish translation of newscasts from four of its stations for distribution via their YouTube channels.
Rafi Mamalian, vice president of the Sinclair Innovation Lab, told me during an interview the station group’s initial intent was to ensure it could reach and warn more people during severe weather events like hurricanes in South Florida.
“Then it dawned on us. There are probably a lot of underserved audiences in a lot of our markets where we could potentially turn on the live broadcast and see if we could expand our audience,” he said.
My ears immediately perked up because I have written in this space before about this sort of thing. Broadcasters will have the opportunity to serve local viewers better, expand their audience and generate more ad revenue thanks to a future marriage of AI language translation and NextGen TV.
According to the U.S. Census Bureau, there are some 68 million people in the United States who speak a language other than English in their homes. While Spanish-language speakers by far outnumber speakers of other foreign languages in the United States, the number of Chinese, Tagalog, Vietnamese and Arabic speakers is significant. For instance, the bureau puts the number of Chinese speakers at nearly 3.5 million.
According to the bureau, about a third of those Chinese-speaking households are designated as “limited English-speaking,” meaning a household where no one over the age of 14 speaks only English or English very well.
In a 3.0 future with support for personalized audio and multiple channels, why wouldn’t a broadcaster in San Francisco, for example, simulcast an AI-generated Chinese-language audio translation of its English-language newscasts?
From where I sit, deploying AI foreign language translation and marrying it up with local newscasts transmitted as 3.0 signals will be a win for viewers, a win for broadcasters and, thanks to helping more people be better informed, a win for society at large.
