Artificial intelligence was arguably the hottest topic at IBC2017, but what can broadcasters hope to achieve by investing in the technology?
As with any transformational technology the first issue to address when considering adoption can often be found in identifying the urgent from a long list of requirements.
In broadcasting - which brings a wide range of potential use cases for cognitive computing - this can mean beginning with a very long list.
Many traditional broadcasters have a desire to achieve efficiency gains. Others want to dive into areas as data, understanding more about the viewer and using data to inform the commissioning of content.
There is a standard adoption curve for any new technology, the difference in broadcast being that decision times are shortening and the adoption curve is moving through the phases at an accelerated rate.
Carrie Lomas, IBM Cognitive Solutions and Internet of Things Director, believes that when it comes to understanding and engaging with cognitive computing and artificial intelligence the majority of so-called traditional media companies have moved beyond the proof of concept phase.
“We are looking at scaling up in media and broadcast and not just looking at efficiency” - Carrie Lomas
As Director of Cognitive Solutions and Internet of Things Carrie Lomas, works within the Global Services division of IBM, where she runs a large team of consultants helping clients in the media space. In terms of cognitive computing, her objective, she says, is “help someone make a decision rather than make it for them.”
They want to see themselves as part of the new way of working and as you’d expect they are trying to figure out how. Broadcasting and media is populated by many stakeholders, each bringing their own agenda. So a key question for suppliers is who can you get into the room when discussing AI and cognitive with a broadcaster.
“When we engage, who attends the meetings is a tell-tale sign, that’s how I know we are looking at scaling up in media and broadcast and not just looking at efficiency,” says Lomas.
“In one broadcaster we had the head of programming, we had the head of advertising, we had the scheduler. It was really broad, and really exciting to hear the discussions over the adoption. People understand and see what it can do for them,” says Lomas.
For example, some broadcasters want to know how to create excitement for commissions or if they have to execute a big launch for a series already commissioned, how to understand the client base by age group, gender and so on to more accurately target viewers.
Others want to understand how they can get into areas such as programmatic and targeted advertising. “When we discuss programmatic, we talk a lot about using Watson IBM’s to gain personality insights,” she says.
Use Case One – Live News
Efficiency gains are not being ignored. One use case for live news is how to use cognitive computing to explore an archive. If you can tag an entire archive it improves access to the material.
If you are producing live TV, ensuring instant access to archive material is vital, especially for news broadcasters in times of political upheaval.
Use Case Two – Sport
Sport and cognitive is another big use case. IBM’s engagement with content ranges from owning the Weather Channel to managing the data for tennis grand slams including Wimbledon. Its use of IBM Watson, enables broadcasters to access data on matches going back 40 years and broadcast deep stats on players going back over ten years which can be represented using visualisation tools.
The power is in applying multiple APIs (application programming interfaces.) Lots of APIs gives better information.
The discussions around cognitive are that broad. It has moved beyond to proof to acceptance of what it can do.
“What we would normally do- the engagement process – is we would typically sit down and look at pain points and opportunities. We have gone through the fear cycle. As things speed up new technology is adopted at a much quicker pace. The work typically involves design thinking. Some want to move rapidly and jump as first movers. Some are very clear about wanting to understand where they see the value.”
Some want a journey mapped out not a big bang. For some, the question is about achieving a return on investment within a year. For others, such as a when the 2017 General Election was called, the urgency was “What can I do tomorrow?” It is a mix of the strategic and the tactical.
In news media the election was a catalyst. Taking that as a starting point broadcasters are already adding different layers of AI technology and the more they practice, the more understanding they are gaining.
The advice is to think about where you want to apply AI. Are you thinking strategically or tactically? Think about the application. There is huge value in first thinking about what you have today.
And if it in the end it all comes down to eyeballs and how they are measured, this presents the perfect use case for AI. How success, in terms of viewing figures or readers, is measured in many different ways. How and what has been measured hasn’t really changed for years. But it is changing now.
Says Lomas: “If we understand more about people – and more about their choices then you bring your client to you because you know when and how you need to bring the product to them. That’s all part of how understanding decisions and that for something to work successfully it is much better to understand if and where it has value. It is that understanding that AI brings in media and broadcasting. How to deliver it, who to deliver it to, and when.”
Read more Artificial intelligence in broadcasting