EVENT | AI in Broadcasting

In recent years, Artificial Intelligence has been rapidly developing and has began its transformation of industries. While the likes of manufacturing are quick to take up the AI benefits quickly, the media and entertainment sectors are not far behind. Lumina attended the AI in Broadcasting event produced by the Royal Television Society. It proved to be a fascinating into the near-future and how AI can be adopted by broadcasters and channel networks. We’ve taken a look at how AI will influence the future workplace (click here to read), now we take a look at how the industry will change.

Hosted at the ITV London Studios, on one of the final few days of operations, the panel was made up by Ian Whitfield (Virtual AI), Doug Clark (IBM), George Wright (BBC R&D) and Cassian Harrison (BBC Four). The panel was chaired by Enders Analysis’ Andrew McIntosh (Head of TV Analysis).

Each panellist discussed the benefits of AI in Broadcasting, showcasing some of the time consuming tasks that robots could be programmed to take on. Ian Whitfield, former Director of Technology at ITV, began by putting London at the forefront of this emerging technology. London is home to Google’s Deepmind project, which is receiving significant investment, but the tech giant recognises the importance of remaining in London. It is home to several major industries, the financial sector and media sectors, two areas that stand to gain a lot from incoming AI technology. We have already seen the rise of the Fintech sector.

Of the major AI projects by the tech giants, its arguable that IBM’s Watson is the most known and versatile across different tasks. Doug Clark took to the stage to demonstrate the possibilities with Watson, including metadata improvements, highlight reel editing and captioning. The latter takes just 7 minutes for a 21 minute episode, an impressive feat. Watson has also been used to edit together a trailer for Sci-Fi movie Morgan (2017) – click here to watch the trailer.

Yet, while this is an impressive and quick process, AI continues to require human input to check and clarify the machine has performed correctly. With the case of captioning, it would surely still require a human being to continue the process of watching entire episodes to ensure the subtitles match the speech.

AI is also receiving criticism for its technology potentially wiping out millions of jobs worldwide – somewhat justified when an AI computer can edit an entire trailer together. While most significant jumps in technology have resulted in new jobs to cover most losses, yet many remain sceptical that this technology won’t follow the pattern.

For broadcasters, AI clearly offers them a solution for the major networks with large archives, as BBC’s George Wright and Cassian Harrison explained. With BBC Four’s curated scheduling approach, it is important to know everything the archive holds that could relate to a particular subject, therefore Cassian explained how he approached George for an scheduling and archive solution, which resulted in the AI. There were several key questions that the AI had to prove:

  • Could the AI capture the spirit of BBC Four?
  • Analysing beyond just full programmes e.g. content
  • How quickly can AI operate – can it analyse real-time audience feedback?

As content discovery becomes increasingly difficult with the shear amount of content now available, AI in broadcasting will be more crucial in connecting consumers with the right content for them and recommending new programmes and shows.

Our thanks goes out to the RTS London for an excellent event at ITV London Studios. As the ITV moves away from its Southbank home to White City (Broadcast Article), the event was rounded off with a celebratory drink. As for AI, it looks set to increase its involvement in the broadcasting sector, relieving the most time-consuming tasks.

AI in Broadcasting