The National Association of Broadcasters Conference, April 15-19 in Las Vegas, traditionally focuses on the infrastructure and technology used to deliver content to entertainment-hungry audiences. I’m here. This year was no exception, but it was the first time the emphasis was on artificial intelligence.
Fueled by the tech industry’s interest in AI platforms like ChatGPT, multiple technology and infrastructure vendors at the National Association of Broadcasters (NAB) conference discussed how AI can improve efficiency and potentially bring new innovations. I tried to explain what business model we are offering to the broadcast industry.
In discussions with exhibitors at the show, a key discussion on AI focused on how broadcasters, pay-TV service providers and streaming video platforms have used AI over the years, and how the broadcasting industry has We stressed that we didn’t exactly jump on the AI bandwagon at the last minute. .
Specific discussions about AI platforms and capabilities used in industry typically fall into three AI segments: data analytics, machine learning, and generative AI.
data analysis
Both NAB content creators and infrastructure vendors noted that the industry has been using data analytics for years to generate near real-time insights into audience behavior. Use data tracking to identify who is watching what and provide instant feedback to broadcasters and content creators. Companies such as Netflix and Hulu have mastered the use of data analytics to provide audiences with recommendation engine-based program suggestions. Traditional terrestrial broadcasters and cable channels also use analytics to measure the success of their advertising campaigns.
Also important is the ability that data analytics can provide in tracking content consumption outside traditional broadcast environments, such as content viewed on mobile devices and social media platforms.
machine learning
Machine learning in the broadcast industry seems to be defined in several ways, but the most common definition revolves around the use of detailed mathematical models to digest and learn from large datasets. doing. This learning capability can be used to create content ranging from animated cartoons to virtual environments used in games to creating voiceprints used by content creators to replicate voice. Interesting discussions about voiceprints at NAB include using AI to recreate the voices of deceased actors and celebrities, and using voiceprints to translate speech and dialogue within programs into multiple languages. was
During a keynote speech, one executive said: Veritone Co., Ltd., a company that develops enterprise AI solutions, explained how it is providing its AI technology to soccer star Cristiano Ronaldo. Using a synthetic voice platform, Ronaldo can post updates on his personal website in multiple languages, including Arabic, Hindi and Chinese.
Generation AI
While there is a lot of talk about large-scale language models developed by companies such as Open AI LLC and Alphabet Inc., generative AI and its potential in the broadcast industry became a hot topic at NAB. Generative AI can be defined as an AI platform that creates or generates new content on its own, primarily based on mining existing patterns and datasets. This generated content can range from text to images to software code, but also includes video and audio.
Generative AI is less widely used in the broadcast industry than machine learning and data analytics, but its potential is seen as significant.
One use case for generative AI is focused on making existing broadcast content more accessible by compressing existing video clips into bite-sized chunks of content. In this example, a generative AI platform can take an hour-long video of him, such as a TV documentary or news program, and generate a 60-second summary that can be distributed online or via social media.
Another potential use case for content creators is the ability of generative AI to find and produce specific videos that can be integrated into the production of new content. An example discussed at NAB was a 30-second clip of him covering the previous night’s baseball scores, and a sports show making the action of the game recorded as the background of the clip. Using a generative AI platform, the producer said, “Show me yesterday’s baseball highlights,” and the platform searched existing online baseball footage and created a suitable series of video clips.
A real-world example of the use of generative AI in the US broadcast industry was announced in early May. Gray Television Co., Ltd., the second largest TV station in the country. Gray has struck a deal with Waymark, a Detroit-based AI solutions developer, to allow Gray’s local TV station to use his generative AI video production platform.
Waymark’s AI solution enables local businesses to produce narrated TV commercials in less than five minutes. The companies said Gray intends to eventually use the platform on all 113 local TV stations.
not a magic wand
AI in the broadcast industry was one of NAB’s major themes, but some industry voices at the show were less positive about the potential impact of AI in the near future. A senior official Cina Media Co., Ltd.The company, a leading video technology provider to the broadcast industry, warned that while AI could offer great potential for some broadcast applications, it is “not a magic wand.”
The executive said AI is typically deployed on a case-by-case basis, whether it is data analysis tools, machine learning capabilities or generative AI platforms. The executive also said that most broadcast and video AI applications take time to improve and become useful, so for those seeking AI that is rapidly transforming the industry, the executive’s advice is to ” Be patient,” he said.
Technology is a regular feature of Kagan, the media research group within S&P Global Market Intelligence’s TMT service, providing exclusive research and commentary.
This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, an independent management unit of S&P Global.
