NAB 2024: Assessing the AI Revolution in Entertainment
The signal-to-noise ratio in today’s relentless AI buzz is far from optimal—particularly at NAB 2024, where the Everything AI vibe was off the chart from the moment the show began—but one session that cut through the noise and swapped hype for refreshing insight and candor was titled The AI Revolution in Entertainment: One Year On… Moderated by ISO/TC 36 Cinematography Chair Andy Maltz and featuring three senior practitioners from far-flung corners of the media world—former Sony Technology Development CTO Bill Baggelaar, Intel Head of M & E Partnerships Rick Hack, FOX CTO Melody Hildebrandt, and VSI CTO Scott Rose, the session explored AI’s recent successes and failures and where it’s made inroads or failed to do so in current media pipelines.
Comparing the constant gushing of daily AI-related news stories and hype to “drinking from a fire hose,” Maltz reminded attendees that major technology changes happen frequently—”we’ve all been through them”—and posed the question, “Is this really a revolution?”
Not Ready for Revolution
Baggelaar’s answer was, at best, a qualified “yes,” and one mostly qualified by the state of tools that are still relatively primitive compared to what they’re likely to become. Acknowledging that AI represents disruptive “tech that’s going to change the way we work,” Baggelaar asserted that the reason the reason “the disruption is not as daunting today” is that for all of generative AI tools’ remarkable capabilities, it still take considerable work from humans to realize their potential.
Tools like OpenAI Sora that are designed for content creators have not reached the stage where you “tell it what you want and get what you like. There’s still a creative process. You can create great stuff, but it’s not easy… There are future possibilities, but it’s not taking over any content creators’ jobs today.”
Not Necessarily the News
Describing herself as a “super AI optimist” who believes “there’s no more exciting time to be involved in media,” FOX’s Hildebrandt also cautioned that with great opportunity also comes great responsibility, particularly in her field, in light of the “commercialization of news.”
“Content creators should be pretty principled” when it comes to how AI models are created and used, transparency about what is real and what is not, and “binding each piece of content to its creator so that consumers can determine that it’s from a source that they trust.” Predicting that “In five years, 80% of the internet will be AI-generated,” Hildebrandt maintained that content creators should play an active role in determining how large-language models are created “rather than waiting for big tech to do it for us.”
She added, “We’re very much in the Napster age in how LLMs are connecting content with its source. We need to move into the Apple Music age. How can we put guardrails around our content? How can we use LLMs with confidence in the enterprise? What inputs went into a model, and how were they changed?”
Localize It
VSI CTO Scott argued that one area of great potential for AI is automating, accelerating, and enhancing the localization of content distributed globally. For AI, he contended, localization is “low-hanging fruit.”
One area “where LLMs have an increasing impact,” he said, “is the ability to accurately translate content and allow for idioms,” or regionally specific colloquial expressions rather than straightforward translations that lose their impact through formality.
Beyond text-to-speech and speech-to-speech, another AI-enabled aspect of localization is the ability to change images, which impacts localized content’s ability to comply with local cultural issues or regulations by altering content that is considered unacceptable or offensive. Beyond the removal of violent or sexual content, this may include identifying smoking or drinking, or even changing “a bottle of beer to a bottle of water.” AI gives global content distributors reaching into different markets the ability to “solve problems qualitatively.”
The Content Value Spectrum
Scott also noted that the ability to dub new audio tracks with AI makes it possible to effectively “reanimate someone who has passed away,” citing the example of Andy Warhol’s AI-generated voice reading his own diary in the Netflix documentary The Andy Warhol Diaries (2022).
But as with the re-generation of commercialized news content often detached from its point of origin, Scott noted that leveraging AI’s remarkable ability for automated dubbing also means assessing the need for human intervention, depending on the project and the stakes. Predictably, he said, “When there is no human in the loop comes the risk of errors.”
So, when is human QC absolutely necessary, and are there circumstances where AI is good enough? Scott shared the content value spectrum shown below, dividing it into four tiers, asking the question, “Where is AI appropriate?” And how will that change over time?
“When you use AI,” Scott said, “you have to know your content. AI has huge value for automating audio descriptions in games. The opportunity is high, but it’s also more risky in the higher tiers. As you go up the tiers, there’s a lot more involvement by humans.”
That these prescriptions are likely to shift over time doesn’t necessarily mean reduced roles for humans; it more likely means different roles. “And as we start to expose these technologies in Actors’ unions,” Scott said, “we need to explain how their role is going to change.”
Related Articles
On Monday, May 20, at Streaming Media NYC, two lively debates on the topic of AI in Media and Entertainment showcased leading industry figures. The first, "AI: Boon Or Bubble?" examined the ways that AI has entered every sector of the media, business, and technology worlds, and it explored how much of AI's insurgency is a fad, and how much of it is the actual future of streaming. The second, "The Real Value Of Artificial Intelligence," explored the fearmongering around the idea that "AI is coming for our jobs," and focused on how AI/ML might make workflows more efficient and effective, grow or refine reach, and help to boost bottom lines.
20 May 2024
In this interview from the Blackmagic Design booth at NAB 2024, Blackmagic Design's Bob Caniglia and Streaming Media's Shawn Lam discuss how Blackmagic is enabling producers to convert 4K and HD signals to SMPTE 2110 so they can move content across IP networks, with their new open-source 2110 IP codec and new 10 gig port-equipped Blackmagic Design cameras that support it like the PYXIS 6K and the URSA Cine 12K.
29 Apr 2024
Among the key features of vMix 27 are Zoom integration, enabling remote streaming producers to bring in (theoretically) an unlimited number of remote guests, vMix Senior Systems Engineer Heath Barker reports in this interview with Streaming Media's Shawn Lam in the vMix booth at NAB. Barker also does a quick hands-on demo of how the feature works.
29 Apr 2024
Cory Treffiletti, CMO of Rembrand, discusses the company's innovative use of programmatic generative AI for virtual product placements, along with how speedily AI is developing within the advertising industry and the ways it is upending the traditional media buying process.
29 Apr 2024
Any NAB report is like the story of the blind man and the elephant: what you experience is what you touch, representing a fraction of the whole and perhaps not even a good sample. That being said, here's what I touched during the show. Many of these experiences are accompanied by video that I shot of the interviews.
23 Apr 2024
On the show floor at NAB 2024, Shawn Lam of Streaming Media and SLV Live interviews Atomos CEO Jeromy Young about the new Atomos Ninja Phone, which turns an iPhone 15 Pro or Pro Max into a 1600nit, 10-bit, 2,000,000:1 contrast ratio, 460ppi, HDR OLED, ProRes monitor-recorder for any pro HDMI camera.
22 Apr 2024
Exciting new and (mostly) AI-driven tools and services from NAB 2024 that very specific problems, from shooting great iPhone footage to automatically creating short clips to providing live low-latency translation and captioning to creating customized radio programming to building purpose-driven social communities.
19 Apr 2024
As broadcasters and content providers head to next week's NAB conference in Las Vegas, a quiet transformation is unfolding within the world of content and, more specifically, content delivery. It isn't just content evolving, but how it is consumed and delivered. That's why connectivity is a major theme at this year's event.
11 Apr 2024
A lot of products tagged as "AI-powered" were doing exactly the same thing before AI. But labeling their familiar features as "powered-by-AI" makes the same tech seem fresh and innovative.
11 Apr 2024