AI at 4.0 forefront
As Thailand pushes forward to the digital economy era, ramping up the use of artificial intelligence will serve not only to transform the world of business, but how people live. By Suchit Leesa-nguansuk
Artificial Intelligence (AI) in Thailand is moving forward at full steam, thanks to the development of smart algorithms that can replicate, assist and enhance human tasks in predictable and automatic ways.
Jarit Sidhu, research manager at IDC Asia Pacific, a leading global research firm, says: "Thailand is moving forward at the right pace to integrate emerging technologies like AI into business and daily life."
There are two types of AI: applied and general. Applied AI systems are designed for particular areas like trading or autonomous cars. General artificial intelligence systems, such as IBM's Watson, on the other hand, are engineered to mirror human capacities and can potentially handle any task.
The Watson platform is now powered by 30 underlying cognitive technologies, including natural language processing, machine learning and deep learning. These capabilities are designed to solve a wide range of practical problems, boost productivity and foster new discoveries across many industries, says Kittipong Asawapichayon, country manager for the software group of IBM Thailand.
He says cognitive technology like Watson will enhance professionals' expertise and support decision-making by finding the most relevant and accurate information millions of times faster than its human counterpart.
IBM is working with governments, corporations and startups to infuse Watson into areas including customer engagement, fraud protection, cybersecurity, knowledge driven expertise, operations and business processes, healthcare, retail, banking, insurance and telecommunications.
For example, IBM and Visa are collaborating to embed payments and commerce into any device, from a watch or a ring to an appliance or a car. Watson's Internet of Things (IoT) platforms will help companies operate secure payment methods across their entire product lines through the Visa Token Service, a security technology that replaces sensitive payment account information found on payment cards with a unique digital identifier.
Watson has the potential to operate in practically any sector and any environment. For instance, it can alert a driver when the car's warranty or certification is about to expire or if specific car parts need to be replaced.
Its economic impact on retail could also be significant. Watson's order management system can independently go through the tedious process of checking inventories and identify the quickest way of fulfilling an order.
In healthcare, Mr Kittipong says Watson can provide physicians with evidence to support treatment decisions. Bangkok's Bumrungrad hospital is already employing the system to help oncologists plough through the massive volume of new cancer research published each year
In terms of cybersecurity, IBM is powering the industry's cognitive security operations centres. "Watson can help security analysts parse thousands of natural language research reports that have never before been accessible to modern security tools," says Mr Kittipong.
Soranun Jiwasurat, chief of information security at the Electronic Transactions Development Agency, says at this stage, machine learning technology -- a type of AI -- can only solve simple problems, such as matching threat patterns to detect malware, email phishing and spam.
Watson's crucial feature is that it is customisable. Software developers can integrate their own data to train Watson on a specific domain and customise API models to speed up the development of cognitive apps.
Microsoft is competing head-to-head with IBM to help developers bring AI technology to life. Its Azure Machine Learning is a managed cloud service for predictive analytics solutions, while its Bot Framework allows developers to create intelligent bots that can interact with users naturally through various channels.
Furthermore, its cognitive services platform offer developers the ability to integrate AI capabilities into their applications, including APIs for intelligent vision, speech, language, knowledge, and search functionalities.
But AI is not confined to the research rooms of high-tech companies. Chatbots -- text or voice-based digital assistants like Apple's Siri, or Google's Now -- are AI applications that are already part of our daily lives, says Mr Jarit.
Chatbots can also provide business and government a more efficient alternative to traditional customer service channels.
Siriporn Pajharawat, director for the Developer Platform Evangelism group of Microsoft Thailand, says chatbots represent a highly interesting use of AI. Local developers such as Betimes Solutions have used Microsoft technologies to develop bots that can receive input and provide responses in Thai.
Takeshi Idezawa, chief executive of Line Corporation, says the company is developing its own AI platform Clova (cloud virtual assistant), which is offered in a variety of languages.
"Thailand is our second most important market and has always been a priority in terms of new products and user feedback. But we are still working on the exact timing of other roll-outs for Clova after Japan and Korea," says Mr Idezawa.
Ariya Banomyong, managing director of LINE Thailand, says chatbot can help local businesses and startups by making services available within LINE, without the need to download any other apps.
LINE also owns a chatbot translation service and operates a chatbot in LINE Finance's official account.
Among the local brands that have already integrated chatbots into their services are Uber, Wongnai, Citibank, Lazada, Krungthai AXA, Maybank, Shell, Unilever and FWD.
AI-connected homes could soon be as ubiquitous as chatbots. Eugene Yoo, general manager of the global business development cloud centre of LG Electronics, says LG has integrated its robot technology with Amazon's Alexa voice recognition to introduce Hub Robot, a digital assistant that can control and complete household tasks.
With simple verbal prompts such as "Turn on the air conditioner" or "Change the dryer routine," the connected LG appliance will automatically execute the assigned task.
"We will introduce such technology this year in the US and Korea, and then expand to other countries," says Mr Yoo.
Ms Siriporn says many independent software vendors have taken advantage of cloud-based machine learning platforms to develop solutions with predictive analytics capabilities for businesses.
"The model that is currently most prevalent is the use of machine learning to process data into predictions, such as analysing customer data to determine likely future purchases," says Ms Siriporn. "We also expect to see industrial applications where AI is used in combination with the IoT to enable predictive maintenance for machinery."
Mr Jarit says the banking, automotive, retail, telecom and aviation sectors will probably be among the earliest adopters of AI in the country.
Thanachart Numnonda, president of the Association of Thai ICT Industry, says there is an enormous potential for AI in fintech.
Perhaps the largest stumbling block for the development of AI is the lack of technical expertise. If it is to remain competitive in this emerging sector, Thailand will need to cultivate a data science and software development savvy local workforce, and redouble its effort to recruit talent from abroad through special visa permits.
One of the key features of LINE Finance service is a chatbot that replies users' questions and makes stock recommendations. Suchit Lee-sanguansuk