The story of 2024 has undoubtedly been the impact that generative AI has had at all levels of the industry - and that only looks set to accelerate as we head into 2025.
If there is one thing that will link 2024 and 2025 together it is the unstoppable rise of generative AI and its impact at all points of the video industry. It is now just over two years ago that the first Large Language Models (LLMs) were publicly released, and the consequences since have been extraordinary.
A lot of the effort that characterizes the past year across the industry has been about investigating LLMs, assessing what they can and cannot do, and seeing how they fit into the business plans. Or in other words - separating the opportunity from the hype. You can see the attention to detail that is required in our recent post How predicting streaming loads can help broadcasters and operators slash costs. AI, and generative AI in particular, is a complex subject with many nuances that requires deep-level expertise throughout a company to cost-effectively extract the most benefit from the technology.
There are other subsequent trends though that are also emerging as important. Immersive video might have had a slow start in the market, but the capabilities of the technology are improving at mass market price points all the time. And the consumer-facing technologies that AI is helping accelerate, particularly in the realm of personalized content, will start to affect the ways that content is created.
We asked two of VO’s leading experts who are at the front line of assessing and understanding new technology’s impact on the industry their thoughts. Einat Kahana is VP of Product Management and Dror Mangel is VP TV Products & Solutions. This is what they said.
Q: What has been the most significant industry development in 2024?
Einat Kahana: During 2024, the usage of genAI have been demonstrated through various use cases in the video industry. One of the main ones has been the automation of content-related processes.
- Content Repurposing: the creation of different versions of content for various platforms and formats (e.g., creating short-form social media clips from longer videos, generating trailers)
- Automated Video Editing: analyzing raw footage, identifying key moments, and automatically generating highlight reels, which significantly speeds up the editing process.
- AI-powered dubbing and subtitling: which makes localization faster and more cost-effective
Dror Mangel: 2024 has been the year of early adoption of generative AI in the media and entertainment industry. This technology is now being leveraged to create personalized content, automate production workflows, and enhance viewer engagement. Generative AI’s ability to produce tailored promotional materials, dynamic storylines, and even localized content has opened up new opportunities for content providers to differentiate themselves. However, its real breakthrough lies in making these capabilities accessible at a cost-effective level.
Q: Following on from that, what do you believe will be the most significant trend in 2025?
EK: The increasing adoption of AI-powered features that add value to consumers and address their needs will be a key trend, including AI-driven content personalization and content creation. A rise in interactive TV shows and movies will reflect the growing demand for immersive and engaging content experiences that allows viewers to make choices that influence the storyline. Another important evolving focus for the industry is the need to battle against fake news by verifying the authenticity of content before publishing or broadcasting it, thus preventing the spread of misinformation. This will be another important trend for next year.
DM: 2025 will be the year generative AI moves into mainstream adoption, with companies focusing on cost optimization to scale its use effectively. The integration of hybrid AI infrastructures, balancing cloud elasticity with on-premises systems, will play a critical role in ensuring generative AI solutions remain economically viable. Generative AI will expand to audience-facing real-time applications, such as content discovery and more. Those who master the art of aligning generative AI’s potential with cost-efficient infrastructure will redefine the industry's competitive landscape.
Q: What do you envision the industry will look like in 2030?
EK: The video and TV landscape of 2030 will likely feature hybrid business models, blending subscriptions, advertising, and potentially other innovative approaches. It will see a continued battle against piracy and fake news, AI will be central, driving personalized content recommendations and interactive experiences, while esports will claim a prominent position in the entertainment industry. AI will also improve the optimization of content and enhance the operational efficiency; freeing human resources from automated tasks to undertakingmore strategic ones.
DM: By 2030, generative AI will be a fully integrated part of the media ecosystem, driving personalization and operational efficiency across all aspects of the value chain. The industry will rely on a combination of advanced AI tools and hybrid infrastructures to deliver scalable, viable solutions tailored to diverse audiences. User-generated content (UGC) will also become a significant part of TV, as AI capabilities will enable any user to become a professional content creator. Additionally, generative AI will no longer just create content; it will optimize how it is distributed, monetized, and experienced, focusing on measurable business outcomes rather than flashy, speculative technologies. The emphasis will remain on improving profitability and engagement while maintaining sustainable operations.
AI everywhere: A safe prediction
Making predictions about anything in the video industry is, of course, an often dangerous business. But AI having an increasing influence in all areas and at all levels is about as safe a bet as you can get. There are many challenges still ahead including the issues of guardrails, the potential legal problem of some major players in the business having used copyright material to train AI, difficulties with deepfakes, piracy, and much more. However, the potential is enormous and we’re already moving out of the initial phase of wide scale deployments where it was used primarily as a means of automating existing processes. The next stage is where it gets truly interesting.
Plus there is, of course, always room for other technologies to muscle their way in and become disruptive. It might only be five years till the start of a new decade, but look at all that has happened so far in the 2020s. Innovation is only going to accelerate from here.
