SINGAPORE – Instead of sieving through information or filling in a pre-defined form, people using online platform SupportGoWhere will soon be able to describe their situation and needs in their own words to find a recommendation on suitable support schemes.
Singapore is piloting the use of large language models (LLMs) in the platform’s Support Recommender tool to enhance the delivery of government services, Minister for Communications and Information Josephine Teo said on Thursday.
LLMs are deep-learning algorithms that can recognise, summarise, translate, predict and generate content based on knowledge gained from massive data sets.
SupportGoWhere, launched in 2020 during the early days of the Covid-19 pandemic, is a one-stop portal for Singaporeans to find and apply for social services and government support programmes.
When asked when the feature would be available, the Government Technology Agency (GovTech) said it was conducting internal trials of the new interface with other government agencies.
“The feature will be made publicly available once we have tested, improved and ascertained its ability to recommend the most relevant schemes to citizens based on the information they provide,” a GovTech spokesperson said.
The move to upgrade the platform is part of a push by the Government for greater adoption of artificial intelligence (AI) innovation to develop an initial tranche of “scalable, productivity enhancing, and rapidly deployable” products that can transform the delivery of government services to citizens, Mrs Teo said.
Singapore is also tapping AI to enhance productivity and decision-making in urban planning. The Urban Redevelopment Authority (URA) is using machine learning to translate large amounts of public feedback into insights based on topics of concern and geographical areas.
“This feedback analytics prototype will support urban planners with formulating plans that more effectively address residents’ concerns,” the minister said, adding that the URA is also developing a suite of AI-enabled tools under its Plan.AI strategy.
Mrs Teo, who is also Minister-in-charge of Smart Nation and Cybersecurity, was speaking at tech giant IBM’s Think Singapore 2023 event, held at the Sands Expo and Convention Centre at Marina Bay Sands.
The latest moves ride on the “good momentum” in the adoption of AI across the Government since the National AI Strategy was introduced in 2019, she said.
She noted that as at 2022, 85 per cent of public agencies here had implemented at least one “high-impact” solution using AI to boost productivity, with more exploratory projects under way.
Mrs Teo described AI as a “transformative technology that presents tremendous opportunities and challenges”.
She noted that the technology can provide solutions to the challenges Singapore faces.
For example, AI could be part of the strategy as the country looks to meet the healthcare needs of its ageing population, by acting as a tool for precision medicine, allowing for the prescription of more precise and optimal drug doses based on a patient’s medical data.
AI can also enhance Singapore’s green computing ambitions by helping to optimise energy consumption in data centres, Mrs Teo added.
Singapore takes a pragmatic and balanced approach to AI governance, expanding opportunities for people and businesses while ensuring trust and safety, she said.
Mrs Teo pointed to Singapore’s introduction of a Model AI Governance Framework in 2019 to encourage the responsible adoption and deployment of the technology, as well as the 2022 launch of AI Verify, a governance testing framework and toolkit.
Noting that AI Verify has been made available to the global open-source community, Mrs Teo called on system developers and researchers across industries to contribute to the development of more trustworthy and responsible AI.
