Media and entertainment industry players are coming to this year’s NAB Show with a bit more experience leveraging machine learning and artificial intelligence for a variety of tasks, from content creation and improving media supply chains to personalizing content and analyzing data. They’re also bringing a desire to get past some of the hype to discover how these technologies can solve problems.
“Everybody says they do AI these days, and obviously that’s not true,” said Peter Docherty, founder and chief technology officer at ThinkAnalytics, which has just launched a unified AI-powered platform supporting contextual advertising and content monetization, curation and bundling.
“First, it’s AI; then it’s GenAI [generative AI]. There’s always a new algorithm. I call that the ‘hype cycle.’ For us, it’s really about having the right business capabilities for the business problem or opportunity you’ve got.”
That is not to say broadcasters and other M&E companies aren’t taking advantage of what AI can do. For instance, Sinclair comes to the show having completed an AI trial to generate real-time Spanish-language audio translations of newscasts at four of its local stations for distribution via YouTube. Gray Media is attending with a newly refreshed group-wide AI policy that emphasizes innovative uses of the technology within the bounds of human oversight and control and CBS News and Stations is attending with a newly renewed and expanded contract with Veritone to leverage AI to make its vast content library more accessible and marketable.
“Over the last few years, as more content has been processed and as models have been refined, the level of search accuracy and specificity has increased,” said Sean King, general manager of media and entertainment at Veritone.
“Content creators like CBS can leverage that level of transparency to get into their archives — to their content, frankly, their IP (intellectual property) — offering … operational efficiency in the media supply chain and taking advantage of potential tertiary revenue opportunities,” King said.
Archival Assistance
AI is already helping content owners extend the reach and value of their libraries and to create richer metadata to make archives searchable, said Ian McPherson, head of global M&E business development – Media Supply Chain and Generative AI at Amazon Web Services. At NAB Show he expects to see many examples of how the technology improves media workflows.
“We’ll continue to see generative AI incorporated into tools to automate mundane tasks and assist creatively, especially for companies in the production and postproduction verticals,” he said. “More companies are releasing foundational models that can have meaningful applications in media and entertainment, like Amazon Nova, and we’ll see these types of models increasingly integrated with productivity tools.”
Generative AI algorithms designed to achieve a “deep understanding” of what is happening in video frames offer content owners a path to monetize content shorts cut from footage of longer duration, said Quickplay Vice President of Product Management Naveen Narayanan.
“Shorts are becoming the No. 1 medium for people to find what they are likely to watch from a long-form point of view,” he said. “We are heavily leveraging generative AI in multiple steps … to understand the key moments in long-form content, find them to the millisecond and create short clips that have the potential to go viral and engage users.”
Attracting viewers with relevant shorts is only part of the battle. Content owners also need to serve up the right programs and help viewers quickly determine what they’ll be watching. Here, too, AI is improving workflows for broadcast, FAST channels or simulcast playout, said Matt Smith, chief evangelist at Akta.
“Here’s where AI can help,” Smith said. “You can tell AI to schedule 24 hours of drama shows. Then an operator can remove a show that’s not [a drama] and drop another in. That just saved you 10, 15 or 20 minutes.”
Similarly, AI can automatically generate program descriptions in multiple languages and — even allowing for operator intervention to make corrections — cut the time needed to write descriptions by 80%, he said.
Better Know a Sub
AI is helping streaming distributors tackle other key issues such as subscriber lifecycle management and better understanding the ocean of data available to distributors about their subscribers, said Damien Organ, vice president of product marketing at Cleeng.
“Understanding groups of users on your platform more intelligently at a hyper personalized level is a key dimension of something that machine learning and AI will do better than any other system, without a shadow of a doubt,” he said. “Being able to act intelligently based on understanding different groups of customers is critical.”
Only about 20% of people are confident in their data-literacy skills, Organ noted.
“GenAI is changing how people can work with data, and how they understand the data that their own platform is producing. The ability to tell stories revealed from that data with the help of GenAI is something that’s a real game-changer for the businesspeople who run these platforms,” he said.
NAB Show attendees can expect to see AI developments on several fronts, including service enhancements, such as algorithmic optimization of content delivery network (CDN) traffic, as well as viewer experience-centric enhancements providing real-time content analysis to support interactivity, according to Julien Signes, senior vice president and general manager, video network, at Synamedia.
“You have to scratch a lot of surfaces to see what’s just hype vs. what is real,” he said. “The good news is we are seeing things happening, and I think the future is bright.”
© NAB
