We heard from the BBC, IBM and the nascent AI industry – and all were agreed AI will impact how television is made. Broadly, AI is teaching computers to do what humans do – leading to machine learning – machines that can learn autonomously.

Ian Whitfield, founder of Virtual AI and a former director of technology at ITV, explained how the systems his company were now putting in place could further enhance the automation of many back office processes required to fuel the broadcast of TV programmes. “In channel scheduling, for example, many of the repetitive and time consuming tasks can be done by robots” he said, with an audit trail for rapid monitoring of their decisions.With a low cost of entry, he suggested payback for a straightforward system could be less than a year.

IBM Watson is IBM’s AI brand, and Doug Clark, IBM’s global cognitive solution leader told us Watson is already helping to speed up many of the processes involved in programme production. “We’re analysing text, audio and video to create enhanced metadata and new ways of presenting content, with opportunities to mine and monetize it, and service the content more effectively” he said.

BBBC4 channel editor Cassian Harrison explained a current project with BBC R&D to use machine learning to leverage some of the BBC’s 50 year archive of A-V content. “Finding the best of it is a real challenge – so how can we explore it? What’s possible? Can technology help us capture content that represents the spirit of BBC4 – the kind of content that our audience will love? “ he asked. He also wants the ability to delve deeper into the programmes to find more detail - sequences that could be used in today’s programme needs.

George Wright, head of internet research & future services at BBC R&D explained “We’re working on using AI to make better programmes – the current project will ask – what could an evening of BBC4 chosen by AI look like? And can AI be dynamic enough to respond in real time to audience feedback?”