The government may soon have to choose between allowing massive job losses or letting the private sector be eaten away by regional competitors.
"There is no doubt some technologies have the potential to take human jobs, but we cannot avoid it because other countries are taking the leap. China is already quite ahead, with more robots being sold in China than in the US," said Michael Gryseels, McKinsey's leader of digital practice in Southeast Asia.
In Thailand, technologies could replace close to 55% of jobs, and can handle at least 30% of activities performed in 60% of occupations, said global management consulting firm McKinsey & Co. Jobs concerned with data collection and other routine tasks have the highest risk of displacement.
While job replacement could be widespread, the relative cost of labour compared to technology in the region is stopping many managers from building a case for replacement. The value added and ease of these technologies as well as their regulatory and social acceptance are also important factors in the replacement equation, said McKinsey.
Government officials should not dismiss the increasing potential of massive layoffs, however. The cost of data storage has been falling drastically since the 1980s.
"If you spent US$1,000 on computing power in 2013 you could get the same computing power as the brain of a mouse. If you spend the same amount in 2025 you could get the same computing power as the brain of a human, and if you spend it in 2050 you could get the power of all human brains combined," said Mr Gryseels.
Machines have come close to imitating or superseding humans in skills that were previously thought to exclusive and inimitable, like image or speech recognition.
The utility of traditional artificial intelligence (AI) and machine learning applications is surging as the total amount of data available for processing explodes. "At least 90% of data generated on the planet was generated last year," he said.
This trend in data generation has been fuelled by "simple things like taking pictures and uploading them to the internet or commenting on social media," said Mr Gryseels, "which gives new insight to different facets of consumer lifestyle and the usage of services."
Regional players are catching on. One McKinsey study found only 6% of the region's large companies used the terms advanced analytics, AI, machine learning and Internet of Things (IoT) in 2015 annual reports. Over one in three did so in 2016.
The bulk of innovation in these fields has come from international players, however. Without state support, the gap between these tech giants and Thailand's national champions may widen -- perhaps irreversibly. Companies have invested close to $39 billion in AI in 2016, according to McKinsey, but close to 80% of this investment was done by Alibaba, Google, Amazon, Baidu, and Facebook.
As the benefits of this technology contribute to exponential competitive advantages, falling behind is hardly an option. "There is some risk of monopolistic market structures emerging as early adopters lock in talent, acquire smaller innovators, and capture a disproportionate share of the economic surplus," according to McKinsey.
"In telecommunications, AI can help save 15-30%, and generate a 2-5% revenue upside. In banking, the savings figure is 10-15%, and the revenue upside is 3-5%. In retail, savings are 10-20%, with a 3-6% revenue upside," said Mr Gryseels.
Thailand has already fallen behind Singapore, Malaysia and Vietnam in AI adoption in most industries, with the possible exception of telecom, according to the consulting firm.
Whether and how the government will push a digital transformation that could leave millions without a source of income is still at question.
On the social policy side, the state could establish programmes that redistribute the monetary gains of these technologies -- by offering transfer payments, for example, at least until displaced workers are re-employed or retrained.
Choosing AI over jobs is not a defeat for social welfare. AI is already being used to implement a more inclusive society. Financial institutions can use data from mobile phones to provide insurance or loans to sectors of the population that were previously largely unbanked, such as informal workers.
Alibaba's Ant Financial can make predictions of how creditworthy you are based on the e-commerce transactions you made on its affiliated platforms, for example. "As an insurance company, you can use data from a telecom firm to know if the person you are considering giving motorcycle insurance to drives a lot, drives at night and so on," said Mr Gryseels.
The Thai government could lead the way in creating an open data environment for the private sector allowing cross-industry use of data, allowing small startups to access the same data sets as deep-pocketed competitors.
Data sharing could be key to democratising productivity increases. The productivity levels of midsized and small companies have stagnated in recent years, while those of large corporations have continued to grow, says the World Bank.
A robust open data ecosystem can be an invaluable advantage for local players. In China, PingAn set up an ecosystem where data from over 250 million users can be shared between health, food, banking, housing, auto and insurance firms. In Australia, companies in the automotive, insurance, finance and retail set up a data-sharing network, Data Republic.
Investing in infrastructure is also a possibility for the state. "A lot of data is actually not inside a company, but in cloud infrastructure, and none of the cloud players are present in a big way in Thailand. Alibaba built data centres in Indonesia; we need to figure out how to bring players to the country in order to decrease costs," he said.
Having them closer to home brings down costs as companies spend less on internet connectivity to transfer the data. "We are talking about huge amounts of data here, on the order of petabytes," said Mr Gryseels.