Ken Lorber, Josh Pine and Deeny Kaplan
This month, The Kitchen announced the appointment of Josh Pine as Fractional CFO of the company. The veteran executive—former COO, CRO, and CFO of XL8, the M&E industry’s first and most renowned AI localization platform—was responsible for building that company’s global commercial operations over a three-year period. "We built what is now the de facto industry standard for AI-generated subtitles and captions, and we started getting into synthetic AI dubbing as well," revealed Pine in an exclusive interview with Señal News.
When asked how good AI really is, Pine answered: "It depends. It’s becoming more and more prevalent, but it still requires a lot of human involvement. AI gets you to a certain point, and then you use humans to make it perfect."
A GROWING PROCESS
Pine explained that The Kitchen is starting to blend AI into its operations to create workflow efficiencies. "We’re not going to save 80% in costs on day one—that’s unrealistic. But if we can achieve a 20–30% increase in workflow efficiency as the industry grows, that’s the goal. Our job, as we expand the company, is to use AI to increase our output and improve our workflows to enable growth," he affirmed.
The language services and dubbing (LSD) industry is seeing only about 2% annual growth in the number of linguists entering the profession, while demand for translated content is growing by 30 to 40% annually. “You can’t grow without technology—it’s physically impossible. You’re either faced with massive backlogs, or you use technology to improve efficiency. That’s exactly what we’re doing,” Pine emphasized.
CLIENT QUESTIONS
Pine stressed the importance of addressing clients' questions from an informed and authoritative perspective. "When I speak with potential clients, I try to momentarily separate myself from The Kitchen. I speak to them as an educator and industry expert. Then the client can decide what they want to do. If they choose to use AI, I step back into my role at The Kitchen and execute based on their expectations. But those expectations must be set in advance—and you can’t do that without educating the client, not just about what’s possible within our company, but across the entire industry," he explained.
“There’s also a learning curve,” added Deeny Kaplan, EVP at The Kitchen. “Some clients think you flip a switch and suddenly you’ve got AI subtitles or dubs. We’ve undergone extensive internal training. It takes time. There’s still manual input. People are both afraid of and excited by it."
"AI is a tool in a toolbox—nothing more and nothing less," said Ken Lorber, President/CEO of The Kitchen. "When it makes sense for a program, you use it to the max. When it doesn’t, you don’t," he added. "But you want to be at the forefront—not sitting back to see what others do," Kaplan continued.
Education has long been a core value for The Kitchen. "We make it a priority to educate our clients. We’re there to answer questions. If we don’t have the answer, we’ll find the expert who does," Kaplan emphasized. "That’s what we’re all about. It is a learning curve—no one picks this up overnight."
Pine also highlighted The Kitchen’s commitment to transparency. "There are people in this space trying to mislead clients just to win business. We will never do that. We will always be transparent and educational in our approach. Because in the end, the relationship has to be mutually beneficial," he asserted. Kaplan added that clients prefer to work with The Kitchen rather than go directly to AI companies because they want someone who knows how to use the tools properly—not just provide access to them.
The decision to use an AI tool depends on the context. "If you have a reality TV show where everyone’s screaming and talking over each other and even a human can’t tell who’s speaking, then the machine can’t either," Pine explained. "That kind of content is not a good fit for AI—at least, not today. But scripted dramas or documentaries with clean dialogue and good sound quality? "That’s perfect," he said.
INTERNAL AI USAGE
Pine explained how AI frees up time for staff. "Someone who used to take three days to translate and create captions for one title can now do three titles in three days. The process depends on the language pair and the amount of training data available. The most robust training data is for English to Latin American Spanish, which has about 95.8–96% accuracy—better than most human translators can achieve consistently. A linguist can now process three or four films in a week instead of one. But going from English to Japanese, for example, only has about 65% accuracy. In that case, you don’t want to use the machine," he noted. Time is money, Pine added. "The more efficient you are with your time, the better your return on investment."
"Our goal has always been to stay at the forefront—to be the company that doesn’t just follow change, but drives it," Kaplan concluded. "At the end of the day, we’re a traditional dubbing company. But one that’s committed to adapting as the industry evolves."