Phil Eade
Phil Eade

How AI will accelerate growth in Thailand’s media industry

Tech October 07, 2018 13:52

By Phil Eade
Special to THE NATION WEEKEND

4,080 Viewed

GLOBAL spending on artificial intelligence (AI) will likely exceed US$46 billion (Bt1.5 trillion) in 2020, according to the IDC, a market intelligence corporation. The prediction comes a year after IDC reported that Thailand had the second highest number of organisations adopting AI among Southeast Asian countries. 



AI has been adopted by the Ministry of Information and Communication Technology to implement services and tools that facilitate information sharing and collaboration, along with supporting decision-making and problem-solving. These initiatives have increased efficiency and brought about more personalised services for Thais. 

While AI has been widely employed across industries, there remain many opportunities to be exploited by Thailand’s $14-billion media and entertainment industry. Broadcasters and media firms in Thailand face efficiency challenges in terms of delivering relevant and personalised content to consumers. An Ooyala poll in December 2017 revealed that 7 in 10 Thai media executives view their media operations as only minimally efficient or not efficient at all. In particular, 25 per cent of respondents highlighted the need to automate their entire operational value chain. 

Drawing on the large amount of data generated from the consumption of videos, AI could improve the management and delivery of content as well as automate operations. Beyond that, AI could deliver better insights to help media executives make informed decisions and develop sound strategies for their business. 

Here are three things AI could do for content production. 

Make worthy content

Content producers are in a race to make original and creative content that appeals to consumers. More often than not, this process requires an extensive review of past content to gauge viewers’ interest and avoid duplication of content.

AI makes it easier for the publisher or broadcaster to understand the specific preference and behaviour of consumers by leveraging machine learning. For instance, if user X watches video A then clicks to watch video B, we can infer that the likelihood of user Y also wanting to watch video B is high if they had also watched video A. These patterns feed into a powerful neural network of decision-making that content producers could rely on to create personalised content for consumers. 

Couple this with an AI-powered in-depth calculation, which tells content makers the type of content they should prioritise and focus their production spending on. A decision-maker would be more empowered to make a calculated decision rather than relying on intuition or superficial assessment. Imagine that you have a hunch to proceed with producing a family drama series as the last such series was a success. However, once you’ve factored in the full production costs against your revenue and returns, you might realise that a comedy series would be more profitable. 

Give content a re-birth

As a lot of money is being spent on producing a piece of content, the ability to make the most out of it is key. Localisation of the content to cater to each market’s needs and relevance is one way – and AI could help speed up the process. Tech giants including Microsoft and IBM have AI tools that makes it easier to transcribe or translate audio into other languages, so that you can sell your content to new markets. 

Leveraging existing content is another way to profit. We can almost always hear the same pain point from broadcasters or publishers: “I don’t even know what I have in my archives at this point.” Locating, let alone monetising, relevant footage to supplement a news bulletin or a documentary often becomes a huge challenge for content companies as they sift through the vast amounts of video assets that sit in their archives. With AI, media organisations can index existing content with powerful metadata capture techniques, enabling them to easily identify and repurpose that archived content.

Monitor live events

The challenging issue with live events is that there are a lot of moving parts and if one thing goes wrong, it could cripple the entire event. That also means that monitoring a live event or troubleshooting issues is typically performed in silos, leaving media organisations in a fix as they try to accurately pinpoint and resolve issues. For example, AI could watch out for unexpected spikes in viewership during the live broadcast and alert companies immediately to add server capacity and prevent the event from crashing. This gives production companies the opportunity to monitor and troubleshoot more quickly and efficiently, sometimes even before they happen. 

Never too late for AI

The benefits of AI are exciting. Streamlined workflows, increased efficiencies, and improved monetisation are all potentials that could very well be fulfilled. 

Additionally, with Thailand becoming an increasingly mobile population, only media companies able to deliver content in various modes and formats in a timely manner will gain.

However, AI is still in its nascent stage and still ramping up across sectors. While it will in the foreseeable future become more advanced and integral in our tasks, that does not mean that human participation would be forfeited. Ultimately, AI should play the role of a work-partner able to give insightful answers, but it will still be up to humans to make the right decision.

Phil Eade is senior director for media platform business development, APJ, at Ooyala.