Viet Nam, like many other countries, is in a bid to spur economic innovation with Artificial Intelligence (AI), but experts say the country is starting from at a low point and needs to make a huge effort to catch up with the global trend.

Ki-Ki, the first Viet Nam-made chatbot, was presented at Zalo AI Summit 2018 in late December 2018.— Photo enternews.vn

Such efforts would include focusing investment on training skilled workers and building an open-source database.

Bui Hai Hung, a researcher from Google DeepMind (USA), said the AI advances of the past decade could create comprehensive changes in global industries and services, from healthcare to energy.

He believes the AI laboratories established so far in Viet Nam, such as those started by FPT and Zalo, are a good start.

However, training in the field of AI remains limited due to a lack of funding, so State and business investment is sorely needed.

“Although Viet Nam has paid great attention to AI applications, it still lacks a staff capable of accessing AI research at a global level,” Hung said.

Herve Vu Roussel, head of data engineering at Sentifi, acknowledged the problem.

“A lack of engineers having profound knowledge on machine learning, or data scientists, is one of the challenges for AI development in Viet Nam”, Herve said.

According to a report by Nexus FrontierTech, rubikAI and G&H, most AI companies in Viet Nam are in the first stage of development and have fewer than 10 AI engineers. AI courses are offered at only six universities in the country.

According to a report by TopDev, a recruitment network and ecosystem in mobile and IT, Viet Nam will be short 70,000 to 90,000 information and technology (IT) workers next year despite increasing salaries.

In the past five years, many start-ups have developed projects related to artificial intelligence (AI) and machine learning, and are willing to pay an AI engineer as much as 1,678US$ a month, or 22,000$ a year.
However, the report said many businesses could not find suitable personnel.

Duong Trong Tan, founder and CEO of Agilead Global, a firm providing AI training, coaching and consulting services to individuals, organisations and start-ups in Viet Nam, said this can be addressed by enhancing the workforce for IT in general and AI in particular through a combination of university training and enterprise recruitment.

“In the era of rapid change like today, if a school is not associated with enterprises to improve their training programme and put in new content that businesses need or new technology, they can hardly keep up with the pace of economic changes,” Tan said.

Bui Hai Hung from Google Deepmind said basic training courses for AI would be costly.
“It might need investment and support from the Government and enterprises,” Hung said.

Open-source database

Because AI devices learn how humans think and act through the collection of massive data sets, they work best when more data are available.

Herve Vu Roussel from Sentifi said Viet Nam’s weakness is in data and information.

“When you talk about AI, data comes first,” he said. “But there just isn’t a lot of good data here in Viet Nam.

If you take a look at Google Maps, you’ll see that not all the roads are mapped correctly. AI is a data-driven field, so without good data, we’re at a dead end.”

Charles Ng, Appier’s Vice President in charge of Enterprise AI, agreed, saying that for successful AI adoption, a company’s data must be solid.

“Getting solid data infrastructure in place is still a big challenge that businesses in Viet Nam face,” he said.

However, experts said they are optimistic about Viet Nam’s AI future.

“We have become more serious on this matter; what Viet Nam can do now is to strengthen its programme on data engineering, data mining and databases,” he said.

“But most of all, we have to create a community for AI and nurture it,” he said.

According to a survey on 500 Vietnamese firms conducted in 2018 by Vietnam Report, only 13.6% have invested in AI in production and business; 18.2% are studying the model and 18.2% have plans for investment in the coming two to three years. Almost 50% of the surveyed companies do not have plans for AI investment yet.

(Source: VNS)