At RealScreen Summit, Unscripted Leaders Get Candid About AI, Monetization, and the New Rules of Scale

From metadata to dubbing to fandom building, executives say the future of unscripted is already here and it’s operational.

At this year’s RealScreen Summit in Miami, a panel of unscripted executives made one thing clear. Artificial intelligence is no longer a futuristic talking point in unscripted television. It is infrastructure.

Key Takeaways

• AI is now core infrastructure in unscripted TV, not a future experiment—already shaping distribution, monetization, localization, and decision-making.

• AI is an efficiency engine, not a creative replacement: it removes friction and automates low-value tasks so humans can focus on taste, judgment, and nuance.

• Editors and creatives—not engineers—are driving AI adoption, building internal tools (e.g., “commissioner bots”) to speed up development and feedback loops.

• Metadata is a major competitive advantage: clean, structured metadata is essential for AI-driven discovery, recommendations, monetization, and connected TV performance.

• “Garbage in, garbage out” applies strongly to AI: without high-quality data and context, AI tools deliver limited value.

• AI-powered up-res and catalog processing are unlocking legacy content, dramatically reducing time and cost, while still requiring human quality control.

• AI dubbing is practical, scalable, and cost-effective, enabling rapid global expansion—but success depends on market-specific audience tolerance and ROI analysis.

• One-click globalization is a myth: AI lowers barriers, but localization still demands strategic, region-by-region decisions.

• The industry is shifting from mass audiences to niche fandoms: highly targeted channels drive deeper engagement, longer watch times, and stronger monetization.

• Watch time—not raw views—is the dominant monetization metric, especially across AVOD, FAST, and connected TV platforms.

• Content ownership and well-organized libraries are increasingly valuable, particularly as demand grows for structured video datasets to train AI models.

• AI-driven opportunities favor scale and preparedness: rights holders with clean, documented, technically ready catalogs will unlock long-term value.

• Human oversight remains essential for authenticity, legal integrity, and audience trust—full automation risks alienating viewers.

• Transparency around AI use is critical, especially in sensitive areas like legal recreations and news-adjacent content.

• Bottom line: unscripted content is becoming more global, modular, data-driven, and performance-oriented—and leading companies are already operating this way, not theorizing about it.

Moderated by Wrigley Media Group CEO Joe Livecchi, the conversation brought together Kyle Espinosa (VP, Channels Programming & Data Analytics, Fremantle), Dan Jones (CEO, Little Dot Studios), Clint Stinchcomb (President & CEO, Curiosity Stream), and Rachel Stockman (President, Law&Crime Network) to discuss how AI is actively reshaping distribution, monetization, localization, and creative decision-making across unscripted production.

No one framed AI as a silver bullet. But the panelists were unusually direct about where it is already delivering measurable value and where human judgment remains essential.

AI as an Efficiency Engine, Not a Creative Replacement

For Jones, whose Little Dot Studios operates roughly 1,000 channels across YouTube, FAST (free ad-supported streaming television), and AVOD (advertising-supported video on demand), and acquires tens of thousands of hours of content annually, the biggest shift has been cultural rather than technical. The company now drives more than 11 billion organic views per month.

“For me, it’s been giving the teams the confidence to go and experiment,” Jones said, noting that editors, not engineers, are now building internal tools using Gemini and ChatGPT. One such experiment became a “commissioner bot,” an internal AI tool that predicts executive feedback before a script ever reaches commissioners.

“All the basics are done,” Jones explained. “Then the in-house commissioners get to say, this is what I actually want to spend my time on.”

Across the panel, the message was consistent. AI is not replacing decision-makers. It is clearing friction so humans can focus on nuance, taste, and higher-value judgment.

Metadata Is the Unsexiest Advantage in Unscripted

If there was one unglamorous theme that kept resurfacing, it was metadata (the descriptive data attached to content, such as titles, tags, timing, and context that help platforms understand and surface video).

“AI can only do with what it’s being told,” Espinosa said. “It doesn’t really know what it’s looking at unless you provide context.”

At Fremantle, that context comes from pairing AI-assisted up-res workflows (using AI to upgrade older video from standard definition to HD or higher) with clean, structured metadata across nearly 100 years of unscripted game show content. What once took hundreds of hours per year to process can now be done in months, though quality control remains firmly human-led.

Jones echoed the warning. “Garbage in, garbage out,” he said. “If you haven’t got the data right at the beginning to train your models on, you’re not going to get anywhere.”

In an ecosystem driven by AI-powered search, recommendation engines, and connected TV viewing (streaming via smart TVs and devices rather than mobile or desktop), metadata has become a strategic asset, not clerical overhead.

Dubbing, Localization, and the Myth of One-Click Globalization

AI dubbing (the use of artificial intelligence to translate and replace spoken dialogue in other languages) surfaced as one of the most practical and misunderstood tools in the unscripted stack.

Stockman outlined Law&Crime’s expansion into more than 30 international markets using AI dubbing and localized YouTube channels built specifically for regional audiences. “So far, we’re seeing very good performance,” she said, while emphasizing that audience tolerance varies significantly by region.

Jones put real numbers behind the decision-making. AI dubbing through third-party partners can cost about $1 per minute, or roughly $100 per finished hour once quality control is factored in. In contrast, Little Dot still spends up to $2,000 per hour on traditional human dubbing in certain markets, particularly Germany.

“The watch times are way better, the monetization is way better, and the community you build around it is better,” Jones said.

The takeaway was clear. AI lowers the barrier to global distribution, but it does not eliminate the need for market-by-market return-on-investment analysis.

From Audiences to Fandoms

Several panelists reframed success not around reach, but around intensity.

“Audiences are passive. Fandoms are active,” Espinosa said.

Jones described how Little Dot moved from a single history channel to 17 niche history-focused AVOD brands, each serving a specific interest area such as ancient history or World War II. These super-fan audiences engage more deeply, watch longer, and effectively market the content themselves.

With billions of monthly data points flowing in, AI’s role is less about content creation and more about pattern recognition, helping teams identify what audiences love, reject, and actively champion.

The New Economics of Content Ownership

Stinchcomb placed today’s AI moment in historical context, comparing it to cable expansion in the 1990s and the emergence of YouTube in 2005 and Netflix in 2007.

“If you control rights to a fair bit of content, you’re going to do pretty well,” he said.

While he believes text licensing for AI training may have peaked, Stinchcomb sees accelerating demand for video, particularly highly structured datasets used to train AI models. “As the models get smarter, they need more,” he said.

For producers, the implication is clear. Libraries that are well-organized, well-documented, and technically ready will continue to unlock value far beyond their original release windows.

The Human Loop Still Matters

Despite the enthusiasm, no one advocated full automation.

Stockman detailed the legal and editorial scrutiny Law&Crime applied when using AI-generated courtroom recreations (computer-generated visual reenactments based on real transcripts) during high-profile cases like the recent Sean “P Diddy” Combs trial, where cameras were not permitted. The network relied on official transcripts, consulted multiple attorneys, and was explicit with audiences that the videos were AI generated.

“You want to make sure you’re disclosing what you’re doing,” Stockman said.

Espinosa offered a creative caution. “You can automate everything, but then it feels fake and inauthentic. The viewer turns off.”

In unscripted, authenticity remains the currency. AI can scale operations, but it cannot manufacture trust.

Why This Matters Beyond Production

For companies across the entertainment ecosystem, including music and rights holders, the message from this RealScreen panel was unmistakable. Unscripted content is becoming more global, more data-driven, and more modular. Decisions once guided by instinct are now shaped by metadata, machine learning, and rapid experimentation.

As platforms prioritize watch time over raw views and premium connected TV experiences over sheer volume, every element of production, from picture quality to sound design, is being pulled into a more performance-driven ecosystem.

At RealScreen Summit, the executives shaping that future were not speculating. They were describing what they are already doing.

By the Numbers

• Little Dot Studios operates ~1,000 channels and drives 11B+ organic views per month

• Tens of thousands of hours of unscripted content acquired annually

• Fremantle manages nearly 100 years of unscripted catalog

• Law&Crime expanded into 30+ international markets using AI dubbing

• AI dubbing costs ~$1 per minute (~$100 per finished hour with QC)

• Human dubbing can reach $2,000 per hour in select markets

• Niche strategy expanded one history channel into 17 distinct AVOD brands

• Watch time, not views, emerged as the dominant monetization metric

---