The Briefing: How global governments are using AI right now
By Medha Basu and Nurfilzah Rohaidi
Read how countries around the world are making use of this technology to improve services.
The relationship, currently, is symbiotic. Humans make the policy decisions, AI gives the evidence to speed up the process. But this may change, as roles become redundant as AI increases in its capacity.
Just as concerning is our level of knowledge around AI. As Britain’s Chief Scientist, Sir Mark Walport, recently told us, we’re not entirely sure how it works. “Sometimes artificial intelligence can provide a black box which you don’t quite know the workings of,” he said.
Regardless, we know that it’s here to stay, and will fundamentally change how governments deliver public services. Here are some of the key ways it is already working across Asia.
Healthcare
AI has plenty of uses in healthcare. Hospitals can manage care services and operations better, which translates into lower costs. For an industry facing a manpower crunch, there is a big opportunity for AI to ease the strain on existing staff.
Tan Tock Seng Hospital in Singapore has built an “artificial brain” to manage its 1,700 beds efficiently, so that patients that need one, can get one faster. And robots may one day be taking on the role of a nurse or aide, thanks to the efforts of this hospital’s healthcare robotics research centre.
In fact, technology such as AI will play a central role in an upcoming Singapore hospital - Woodlands Integrated Healthcare Trust - to analyse large amounts of data. Robots will come into play in logistics by delivering food and equipment.
Similarly, South Korea will introduce hospital robots this year to transport items and assist in patient rehabilitation. And a chatbot is helping doctors in China treat patients by getting additional information on symptoms so they can be diagnosed faster.
Agriculture
Smart farming and agriculture is another area that is benefiting from AI. When put together with Internet of Things and big data, AI can improve productivity, make decisions based on data, and help identify the right times to harvest crops or pollinate flowers.
In Australia, advances in robotic vision will mean that there will one day be robot fruit pickers and weeders. These robots will be able to identify and pick fruits in a vine, spray a precise amount of pesticide on weeds, and work through the night.
Japan, on the other hand, is using AI to help develop new food crops faster. Plants are usually bred for flavour, yield and resistance to disease; the ten-year breeding process can be cut down two or three times by AI.
Japan is also the first in the world to have a farm entirely run by robots, proposed as a solution to the labour shortage caused by the country’s ageing population.
Public Safety
When it comes to public safety, AI can help in tangible and meaningful ways. Besides improving productivity, AI can be used to identify hidden patterns in human behaviour, which could reveal harmful intent. Importantly, it does this more accurately than a human being.
AI can help speed up law enforcement and investigations by learning from previous experience, as the head of Interpol recently told us. Based on accumulated experience or data, AI can predict the likelihood of an adverse event happening.
At seaports, AI can help scan and analyse the contents of containers—something Singapore Customs is exploring. Machine learning can then identify objects that may not be allowed through the borders, like weapons.
In the realm of crime investigation, AI can lend a helping hand to ease tedious tasks. Singapore employs a robotic arm to compile photo albums to be used as evidence in court, a task that normally takes investigators hours.
In prisons, AI is helping prison guards detect if inmates get into a fight or try to escape. Malaysia has trialled this system and is now rolling it out across the country.
Meanwhile, AI is also becoming crucial to fighting digital crimes. As criminals now have more sophisticated tools and organisational capabilities, cyber security will be one of the most important areas for governments to use AI, according to the CTO of Microsoft Singapore.
Infrastructure
Everyone hates a train or bus disruption. But transport and infrastructure maintenance is gruelling work.
Singapore is trying to predict train breakdowns before they occur and cutting down on manual labour. It is trialling drones to check on train tracks, tunnels and equipment, and will use AI to predict when certain infrastructure are under too much stress.
Self-driving technology is another way AI is being used. Cities can deploy driverless cars to encourage shared mobility, and fleets can be dynamically deployed as and when they are needed. This can go a long way in managing congestion.
In Australia, Adelaide and Brisbane are planning for driverless vehicles. And Jakarta hopes to introduce driverless cars in five years.
Hong Kong is using AI to schedule all engineering and maintenance works on its metro system.
Similar to train maintenance, Singapore is also looking to use drones and AI to inspect unused state property for signs of cracks, flooding or illegal dumping.
Energy
Across Asia, traditional state-owned utilities and energy providers are facing disruption from new technologies and pressure to keep up with customers’ demands.
Japan’s largest utility provider, TEPCO, is looking at AI for predictive maintenance of its infrastructure. AI will help it preempt any failure and avoid downtime for customers, it hopes.
Meanwhile, a British electricity company is considering the use of AI to make its supply systems more efficient. One possibility is to maximise the use of renewable energies using machine learning to predict peaks in demand.
Procurement and payments
Public finance is a newcomer to this field with developments still in the early phases.
In the future, AI could be used to make procurement decisions, like finding price for a product and evaluating vendors.
Singapore is already trialling machine learning to pick out financial anomalies and fraud in government procurement.
Denmark is looking at AI to decide who should get financial support from the government. This would mean that a machine would process applications for business grants, and welfare payments for citizens.
Japan is planning on using AI to automate “cumbersome” tasks in patent, trademark and design applications, such as literature reviews.
Predictive customer services
AI in the form of virtual assistants and chatbots can improve the user interface of government digital services. This allows to personalise services and also cut response times to citizens’ complaints.
New Zealand is looking at AI interfaces to make governments services more of a conversation and less transactional. This will also allow it to make these services more predictive and personalised. It also plans to use AI for real-time verification of citizens’ digital identity.
Dubai has built its first AI assistant to respond to citizens’ enquiries on electricity and water. Meanwhile, Singapore’s tourism board plans to predict and customise experiences for visitors. One of its first steps has been to build chatbots to answer basic questions from tourists.
Reading all this is enough to make anyone worry about their jobs. But AI is “not a substitute for humans”, Walport said. He believes that the “best possible way” to use artificial intelligence is as an advisor—meaning that we shouldn’t be working against AI. It should be working for us.