AI is transforming the content search and discovery experience, but an ungrounded large language model (LLM) can only answer based on its training data. This, says Gracenote, limits its ability to stay current. In the world of TV, this limitation can have a negative impact on the user experience.
The unit of The Nielsen Company recently asked an “ungrounded” LLM basic questions about 1,300 popular TV episodes across 13 countries.
For nearly half of them, it had no information at all. “Even Stranger Things wasn’t safe,” Gracenote notes of the now-concluded transformative series for Netflix. “When asked about one episode, the model combined details from several episodes.”
Can AI find the right TV episode? “Not reliably,” Gracenote concludes. “That’s why it’s so important for publishers to get episodic content discovery right for audiences.”
While Gracenote’s business is wholly focused on this solution, the data are more than a plug for Nielsen.
“Not only does the amount of episodic content dwarf the number of movies available, audiences search for TV episodes differently than they search for movies,” Gracenote shares. “Viewers rarely know the specific season or episode they want to watch, especially in cases where a show has been around as long as mainstays like NCIS, Grey’s Anatomy and The Simpsons.”
As such, when they search, they use contextual queries like “Show me the Christmas episodes of The Office” and “Which episodes of Friends feature Bruce Willis as a guest star?”
In these instances, LLM responses will usually be incorrect, Gracenote concludes. Why? “They base their responses on probability using season and episode numbers.”
For providers, the poor user experience associated with incorrect responses will compound frustrations over growing content fragmentation. “Episodic content plays a big role in content congestion, as TV episodes account for 85% of the content distributed by the six global SVOD providers tracked in the Gracenote Data Hub,” it notes.
As video catalogs grow, GenAI will become increasingly helpful to viewers looking for something to watch. In just a year-and-a-half, for example, the amount of TV episodes available to audiences has increased by more than 21%. “By offering features like conversational search, personalized program imagery, tailored recommendations and real-time sports highlights, GenAI will transform how viewers engage with video content.”
Meanwhile, Gracenote finds that information about new programs often falls outside of an LLM’s training data, leaving them without the ability to help viewers find what they’re looking for. The Steve Carell-led comedy Rooster, for example, is too new for an LLM released in 2025 to have any information about.
“The end result in this scenario is a poor user experience—one that could cause the viewer to look elsewhere for something to watch,” Gracenote says.
For additional insights regarding the use of ungrounded LLMs in content search and discovery experiences, Gracenote points to its Plot holes in AI report.
