In recent discussions surrounding the media industry, a senior executive at a 24/7 news operation in New York raised significant concerns about the pressures of increasing demand for content against a backdrop of reduced team sizes. The executive highlighted the inefficiencies of current processes, noting that it takes an average of five minutes for a producer to search for a specific shot within extensive media libraries. This translates to roughly an entire workday spent merely gathering clips for a single 10-minute story package, a situation deemed unsustainable, exacerbated by ongoing layoffs.
As technology continues to evolve rapidly, traditional media workflows are undergoing a transformation, particularly with the advent of multimodal AI and generative AI (GenAI). These advanced technologies are notably shrinking content discovery times from eight hours down to a matter of minutes, thereby significantly streamlining the story creation process.
The incorporation of cloud computing has further enhanced accessibility to digitised media libraries, enabling remote access that connects once-isolated media departments. This fusion promotes real-time collaboration across teams, representing a considerable shift in the operational landscape of content creation. The more notable change, however, lies in the implementation of multimodal AI, which functions by mimicking human perception to process diverse data inputs such as video, audio, and text. This advanced AI type vastly improves media indexing by adeptly recognising faces, reading on-screen text, and identifying various visual elements to generate a comprehensive semantic description of video content.
With the ability to search for specific clips rather than entire video files, content producers can gain insights from detailed metadata. This technology proves invaluable, especially during time-sensitive live reporting events like the U.S. elections, where rapid access to compelling shots is paramount. Furthermore, multimodal AI facilitates the creation of niche content packages tailored to various audience interests, potentially enhancing advertising opportunities.
While the media industry may not yet be at a stage where AI can fully produce blockbuster films, generative AI is demonstrating its potential in both pre- and post-production roles. A recent partnership between Lionsgate and Runway aims to develop a model that allows for cinematic video generation, with the studio anticipating significant cost savings by integrating GenAI into their creative processes. This aligns with discussions held at the Tokyo International Film Festival, where industry leaders acknowledged AI's capacity to facilitate considerable savings in production costs, particularly through the reduction of location shoot expenses.
As production companies face mounting pressure to deliver more engaging content with limited resources, multilateral and generative AI allows for the exploration of vast media archives. This technology makes it possible to unearth previously unnoticed footage suitable for repurposing into new formats such as docuseries or behind-the-scenes specials, thereby creating additional revenue streams without the financial burden of expensive shoots.
In the realm of visual effects, notable applications of GenAI are emerging, as evidenced in Tom Hanks’ latest film, where effects startup Metaphysic employed the technology to manipulate the actor's age digitally. This type of work, which traditionally requires extensive human resources and time, is becoming more efficient through AI.
The future of content production is also set to benefit from prompt-driven experiences that increase efficiency in assembling rough cuts. Content producers can specify the narrative they are aiming to build, prompting the AI to automatically sift through media collections for the relevant clips. In addition, AI’s capabilities extend to quality control, allowing editors to use targeted commands to isolate and review specific content elements, thus ensuring compliance with audience standards.
Furthermore, the efficient indexing capabilities of AI grant unprecedented access to archival content, potentially unlocking lucrative revenue opportunities. Licensing footage can command steep prices, and employing AI to index thousands of hours of video archives can vastly reduce the time taken compared to traditional human methods. Advanced models can even ascertain the contents of physical tapes by analysing documentation, allowing for strategic prioritisation during large-scale archive digitisation initiatives.
As the media and entertainment landscape is reshaped by these technological innovations, many legacy systems and workflows face obsolescence. The ongoing advancements in multimodal and generative AI present substantial opportunities for organisations to refine their processes, enabling greater productivity, the discovery of hidden treasures within archival content, and the establishment of new revenue pathways as they move forward in a competitive industry.
Source: Noah Wire Services