Sergey Nivens -

2019: a whirlwind year for AI in APAC

Asia-Pacific is proving to be a magnet for entrepreneurial activity from AI firms, though the escalating US-China trade tensions have put one firm in a tough situation

Like much of Asia in 2019, Malaysia is championing the adoption of artificial intelligence (AI) as part of its overarching digital economy aspirations.

For one, the country’s communications and multimedia minister Gobind Singh Deo wants to propose a national data and AI policy as an important milestone in driving Malaysia’s push towards the fourth industrial revolution.

In addition, the Malaysia Digital Economy Corporation is developing a national AI framework, building on its national big data analytics initiative in 2015 with the aim of fuelling Malaysia’s AI ambition.

No deadlines have been set for either initiative. “I don’t want to put a timeline, because we are not rushing into it. It’s important that we understand data and know how we can use it,” Gobind told reporters during a tech event in Kuala Lumpur in 2019.

Against this backdrop, Computer Weekly reached out to industry players and observers on how AI is currently being perceived and used in Malaysia and elsewhere in Asia.

Sudev Bangah, managing director for IDC ASEAN, described the maturity of AI adoption among local enterprises as “fragmented and opportunistic”, with 65% of enterprises in Asia-Pacific (APAC) either introducing or have executed a pilot AI project in some shape or form.

But there are encouraging signs, according to Bangah: “For instance, while the majority of enterprises can still be said to be moving up the curve in enterprise AI, it is encouraging to see that the distinction between AI functions and data analytics is starting to be clearly made.”

AI hot spots

As a region, APAC is proving to be a magnet for entrepreneurial activity from companies of all sizes.

For example, Malaysia’s Twistcode Technologies, a supercomputer supplier founded in 2006, has rapidly risen as an AI solutions provider across multiple sectors to achieve global recognition.

It is also the only Southeast Asian company to have a supercomputer in the top 500 list of the world’s fastest supercomputers traditionally dominated by the US and China.

Twistcode’s CEO, Nurazam Malim, said the company’s self-assembled supercomputer serves clients in sectors such as financial, defence, oil and gas, healthcare and manufacturing.

Nurazam, who wants to see faster local adoption of AI, pointed out that one of the 2019 highs for the company also proved to be a source of frustration.

“An opportunity to work with a local conglomerate – which has businesses in banking, finance, manufacturing and property – was a little too slow to adapt to and adopt technology changes that the company was envisioning for itself.”

To Skymind AI, 2019 was a whirlwind year, according to its Malaysian director and CEO Shawn Tan. “The escalating tensions between the US and China since early 2019 has put us in a really tough situation.”

But there were bright spots too. Among Skymind’s 2019 highlights was the rare distinction of being selected as a partner to work with the Chinese Academy of Sciences via Neurobionix, an AI research lab that is co-developing the AI framework and infrastructure for China, a global AI powerhouse.

“I have observed that the most interesting adoption of AI is happening in China; you can see AI in every aspect of their lives. They have been living with AI for years, while most of the world is still talking about the concept of AI,” said Tan.

IDC’s Bangah pointed out another milestone: “2019 was a point where digital transformation took a turn into deeper understanding in the region, and a lot of this was brought on by the advent of tangible sectorial use cases that benefited organisations.”.

Bangah added that more discussions on AI took place at the right levels, with many organisations in the region moving up the adoption curve, including those in Southeast Asia.

Avoid the “tech buzz” trap in 2020

Amid brewing interest in AI, Bangah advised enterprises to avoid falling into the “tech buzz” trap. Instead, they should set clear goals prior to adoption, whether it is improving customer engagements, spurring innovation or boosting productivity.

“Examining use cases will also become increasingly important, and perhaps one of the ‘non-tech’ qualities that an organisation or line-of-business should possess is the curiosity to examine how well AI actually enhances their business,” Bangah added.

Only when national AI adoption hits a certain maturity level should a government introduce relevant frameworks and regulations
Nurazam Malim, Twistcode Technologies

On the government’s role in driving AI innovation, Nurazam said governments should encourage their agencies to adopt AI, as well as engage local AI startups to participate in those developments.

“A more collaborative approach, which will utilise funds and grants already in place, will also help better integrate government and industry initiatives in a more balanced manner,” Nurazam said. “Only when national AI adoption hits a certain maturity level should a government introduce relevant frameworks and regulations.”

The way ahead

Looking ahead, Skymind’s Tan singled out the expansion of his company’s R&D team, along with collaboration in open-source projects and ethics standards with governments and enterprises.

Although China is the dominant force for AI adoption in Asia, Tan said Japan, South Korea and Southeast Asian countries are the next frontier for Skymind.

“Asia is emerging as a force that can compete with the largely American-dominated software industry. With this competition, consumers will have more choice in a more robust market in the long term.”

Twistcode’s Nurazam noted that Indonesia will be an emerging market that many AI firms will focus on in 2020 – though he admitted that it would not be an easy market to tackle due to government policies that require local participation in AI projects.

“For Twistcode, we are targeting to be listed in the top 50 world supercomputer rankings in June 2020,” Nurazam said. “On top of our Malaysia datacentre, we are expanding our federated machine learning technology to power multiple AI solutions while adhering to local data privacy regulations.”

IDC’s Bangah said while 2019 was dotted with AI use cases such as automated customer service agents and sales process automation, other areas where AI will be of interest are fraud analysis, preventive maintenance and threat intelligence.

“Innovation, the thirst for intelligence, and rapidly changing customer expectations will drive AI adoption within organisations so they can react quicker to changes around them in the coming years,” he said.

Read more about AI in APAC

Read more on Artificial intelligence, automation and robotics

Data Center
Data Management