We asked seven leading industry personalities for their views on artificial inteligence and how it will impact on the broadcast sector. They are Keith Bedford, General Manager EMEA, WURL; Jonathan Broughton, Head of Research, Plumresearch; Ivan Ljubicic, Managing Director, Pickbox, Mediavision, Mediatranslations and NEM; Anastasia Melnikova, Solutions Architect, Gcore; Katarina Pavlovic, Programme Director Prva TV and B92; Maria Rua Aguete, Executive Director, Technology Fellow, Omdia and Norberts Ositis, CEO, Backscreen.
What is the greatest advantage of AI in our industry?
Keith Bedford, General Manager, EMEA, Wurl
AI is changing the way publishers, streamers, and advertisers connect with viewers on CTV (connected TV). On the one hand, we’re seeing successful use cases of AI-driven performance marketing helping publishers and streamers target relevant audiences most likely to tune into their programming. On the other hand – for brand advertisers – GenAI, especially, is shifting the way we think about performance on CTV. With BrandDiscovery, for example – Wurl’s newest GenAI-based contextual targeting solution – we’re able to analyse the content within the scene closest to the ad break in real time, giving advertisers the ability to achieve contextual targeting at scale that wasn’t possible before AI technology.
Jonathan Broughton, Head of Research, PlumResearch
The greatest advantage of AI in our industry lies in its capacity to boost efficiency and streamline processes. Contrary to fears of AI supplanting human creativity, its primary role often veers towards augmentation rather than replacement. By serving as a suite of tools, AI enhances various production tasks, accelerating workflows, and reducing the specialisation threshold required, thereby democratising access to advanced capabilities.
Ivan Ljubicic, Managing Director, Pickbox, Mediavision, Mediatranslations and NEM
AI has been around for a while, but we just haven’t called it that. Recently the name is connected to both positive and negative connotations but if we learn to use it wisely and with good intentions, it can only help us. The industry faces multiple transformations daily and with the help of AI, we can stay on top of things – from a better user experience to organizations working more effectively. Our goals can be met more easily if we embrace the help of AI and learn how to use it properly. Smart optimisation of web search engines and recommendation systems to creative tools that assist us with research and content creation and organizational programs can only help us plan and navigate through daily tasks.
Anastasia Melnikova, Solutions Architect, Gcore
AI offers our industry numerous advantages, including personalised recommendations, streaming quality optimisation, accessibility, security, expanded reach, and scalability. In my opinion, the most powerful advantage of AI for video streaming right now lies in object and speech recognition. AI makes video content accessible to the deaf or hearing impaired through transcription, and to those who don’t understand the source language through translation. Additionally, AI can help to ensure content safety for different user groups and needs through effective content moderation.
Scalability and the expansion of reach are also crucial benefits. AI enables streaming platforms to handle vast amounts of content and users efficiently, ensuring smooth performance even during peak times. This scalability allows businesses to grow without compromising service quality. AI-driven insights can help companies identify and target new audience segments, tailor content to diverse preferences, and expand their reach to a global audience. So, by leveraging AI, businesses can enhance user experience while achieving significant growth and market penetration.
Katarina Pavlovic, Program Director, Prva TV and B92
We are currently still exploring all the possibilities of AI, watching in real-time its development in various fields, learning about AI while AI is learning about our industry. Today, I see it as a tool which is basically an extension of creators and artists, that will not be able to replace creativity, but it can certainly accelerate creativity and help us realise good ideas faster, especially in content design and production. It can also provide an opportunity for independent and new creators, since they should be able to produce better original content without big teams, high budgets or big studios involvement. We are yet to see how far it can go and what will be the boundaries, if any.
Maria Rua Aguete, Executive Director, Technology Fellow, Omdia
AI brings diverse benefits to the media industry, enhancing efficiency and fostering creativity. It streamlines workflows, reducing costs and provides more room for innovative actions, it will allow people working in media the time to work on higher-value and human facing activities. Some of the current use cases of AI include AI-driven real-time language translation and dubbing that would allow to expand content accessibility globally. Also, in the evolving advertising landscape, AI becomes indispensable for targeted marketing in light of declining third-party cookies. Despite these advancements, challenges such as consumer trust and regulatory concerns persist, highlighting the need for responsible AI practices to build confidence in its application within the media industry.
Content discoverability is a significant issue in the media industry. Users spend an average of 18 minutes searching for content, a challenge exacerbated by the vast number of services and channels, according to OMDIA, there are more than 6,000 subscription streaming video services in the world and more than 2,000 FAST channels.
Norberts Ositis, CEO, Backscreen
I can describe a couple of things where AI is a great helper to media industry.
• Content processing – automatically generated translations, subtitles and voice overs need some review for consistency but still can save a lot of time. Also, generation of metadata, eg. tagging the content, extracting objects, characters and actions in the video, even writing summary descriptions, and generation of thumbnails, posters and trailers.
• Monetisation of the content – detecting automated ad-breaks, AI-supported targeted ad insertion. Recently I heard about interesting concept – product placement in the content, eg. adding billboards to the landscape scenes where the AI can help to choose the best spot.
There are more definitely but these ones can give the most immediate benefits.
How can AI be used to help subscribers navigate content more easily?
Keith Bedford
One of the biggest challenges facing publishers and streamers today is discoverability – finding ways to not only attract viewers to content, but to also entice them to continuously engage. This is where advertising plays an important role in helping publishers and streamers market their content to the right viewers. With AI-powered performance marketing, content distributors can target relevant audiences at scale. The result is a win-win – publishers and streamers successfully attract and retain audiences, while viewers discover new content to love.
Jonathan Broughton
AI has the potential to revolutionize content discovery by addressing two major pain points: choice paralysis and lack of context in recommendation engines. Currently, recommendation systems often fail to capture the emotional and situational context of viewers, leading to this burden being placed on the end user.
Consequently, they face overwhelming choices in vast content libraries. Phycological Studies, (such as those involving jam selection – Columbia 2000), show that beyond a certain threshold, consumers struggle to make effective decisions. AI can alleviate this burden by making contextual and filtering decisions on behalf of users, reducing choice overload, and enhancing the browsing experience.
Ivan Ljubicic
Recommended systems have also been available for a while but lately we are seeing an even more refined way of working – issues with consumers struggling to find the perfect content are taken down to a minimum. Personalised dashboards make the consumer feel comfortable and understood, as well as notifications arriving at the right time with the right information. User behavior can be understood better hence providing an even better content. The crucial part is to encourage the user to interact with the programme – using playlists, providing interface feedback and content ratings will greatly improve user experience and satisfaction.
Anastasia Melnikova
AI offers several functionalities to help users navigate content with ease, including smart search, personalized recommendations, adaptive user interfaces, and enhanced accessibility features.
• Advanced Search: AI uses natural language processing (NLP) to enable advanced search capabilities. Users can search for content using natural language queries, which the AI interprets and delivers accurate results.
• Enhanced Recommendations: Machine learning models analyze viewing habits and preferences to suggest content users are most likely to enjoy. Techniques like matrix factorization, neural collaborative filtering, and reinforcement learning drive these recommendation systems.
• Voice and Visual Interfaces: AI-powered voice assistants and visual interfaces simplify content discovery and interaction. Voice commands streamline the search process, while adaptive interfaces highlight content based on user behavior.
• Context-Aware Suggestions: AI provides context-aware recommendations by considering factors such as the time of day, user activity, and past viewing history to suggest relevant content.
• Accessibility Features: AI can generate subtitles for deaf or hard-of-hearing viewers and translate subtitles for international audiences, making content more accessible to a diverse user base. Search results can reflect a subscriber’s language or accessibility preferences. Gcore Video Streaming offers AI transcription and translation within Gcore AI ASR.
• Content Checks: Videos can be identified and described a video in terms if their content with AI. An AI model can analyse across a set of different categories to provide a detailed content analysis. Subscribers can opt out of results flagged for adult content, alcohol use, or violence. Alternatively, they can focus their search on video content that includes sports, for example, even if the video itself isn’t tagged for sports. Gcore again offers the technology for this with Gcore AI ASR.
Katarina Pavlovic
I see content creation and distribution as an industry heavily dependent on human emotions, so I am sure AI can help immensely, but I don’t think that it will ever function without human involvement. Certainly it can help include more features to the selection algorithm, help discover patterns in subscriber behavior, which should result in more accurate predictions of what subscribers would like, without any actions needed from subscribers, which was method used in the past. However, people are emotional, they change taste and behavior, and I am not completely sure that AI can fully track that.
Maria Rua Aguete
AI could help through personalized recommendation systems that analyse user behaviour, preferences, and contextual data. Additionally, AI enhances content navigation by improving search functionalities and simplifying natural user interfaces, making it easier for subscribers to find relevant content quickly. By leveraging these capabilities, AI not only reduces search time but also enhances user satisfaction, mitigating churn and boosting engagement. This will benefit the media industry by fostering customer loyalty and retention.
Norberts Ositis
AI based content recommendation engines can process a lot of data coming from different sources to proved the most precise recommendations based on viewer habits. Also, natural language processing makes the interaction with the subscriber apps more natural. In the future AI might be able to detect the viewer’s mood from the voice characteristics and suggest the content accordingly.
Would you want to broadcast/watch a movie where a lead actor had been ‘replaced’ through AI?
Keith Bedford
At the pace with which the technology is moving now, we likely won’t even be able to tell in the near future when a lead actor has been replaced by AI on the screen. What will be more interesting, though, will be to see how the advancement of these GenAI tools will ultimately help democratise content creation. We’ll likely see more content creation (and therefore even more content choices) when it comes to what to watch, which will only reinforce the importance of addressing content discoverability and connecting with the right audiences.
Jonathan Broughton
Like in a cartoon? I watch a lot of those already! But in all seriousness, the concept of broadcasting or watching a movie where a lead actor has been ‘replaced’ through AI isn’t as outlandish as it may initially seem. We’ve already seen glimpses of AI’s potential role in entertainment, from fully virtual characters like Gollum in Lord of The Rings, to projects such as ABBA Voyage which made use of AI technology in the rebuilding of famous performances – not an ABBA fan but 10/10 would go again. I don’t feel that AI will truly replace actors on a 1:1 basis any time soon, especially at the premium end, but I think it’s inevitable that due to the inherent efficiencies we will certainly start to see AI generated characters featuring in a prominent way, within the next few years.
Ivan Ljubicic
I’m comfortable with the use of AI in movies and series when it enhances the watching experience – through enhancing visual storytelling, sound editing, virtual reality experiences… But replacement of a human actor with an entirely AI based character will only be a distraction to me – at least until the technology is so refined there won’t be any difference at all.
Anastasia Melnikova
This is a personal perspective, but I always strive to find the positive in AI advancements. AI can create lifelike digital representations of actors using technologies such as deepfake and CGI. Generative adversarial networks (GANs) produce highly realistic images and videos, and by 2030 we can expect to have entire movies and TV series fully generated by AI, not only lead actors. Will I watch them? Of course! How could I not be curious to experience this paradigm shift firsthand?
I see two areas of benefit where AI may be a compelling inclusion, even when it comes to replacing lead characters in movies:
• AI can help create more diverse and inclusive content by analyzing and addressing biases in casting and storytelling.
• AI can open new creative possibilities for filmmakers. For example, AI-generated actors could bring historical figures back to life and offer multilingual dubbing. Directors could experiment with new storytelling techniques and visual styles, pushing moving-image art forms into new territory.
Katarina Pavlovic
I would, both. I think we need to always be open to new technology and participate in its development.
Maria Rua Aguete
From my perspective, transparency is paramount when it comes to AI-generated content. Clearly distinguishing between real actors and AI-generated ones, and ensuring viewers are informed about the use of AI in the production process, is essential. Transparency not only fosters audience understanding but also upholds ethical standards and preserves artistic integrity. By providing accurate information and maintaining transparency, content creators can navigate the ethical complexities of AI-generated content while maintaining viewer trust and engagement.
So yes, providing all the above, I would watch content generated by AI.
Norberts Ositis
Maybe only for curiosity and educational purpose but I doubt that I would enjoy such a lot. While it might be ok for cartoons and perhaps ads, a good movie is something that AI cannot make yet. It just needs a good director’s and good actor’s non-rational personalities.
