S'pore needs environment for quality data sharing to unlock AI innovations: Josephine Teo

Minister for Communications and Information Josephine Teo at the inaugural World AI Cannes Festival on April 14, 2022. PHOTO: MCI

SINGAPORE - Singapore needs to create an environment that allows parties from different sectors to share and create good quality data from diverse sources.

This will help the country unlock more innovative artificial intelligence (AI) applications, said Minister for Communications and Information Josephine Teo on Thursday (April 14).

She was speaking virtually at the inaugural World AI Cannes Festival, which is held in the French city of Cannes and showcases new industry uses of AI.

Mrs Teo, who is also Minister-in-Charge of Smart Nation and Cybersecurity, said that when it comes to data for AI, it is not just the quantity that is important but also the quality: "A smaller, but richer data set can often produce better AI models than a larger, narrower one."

AI models usually need reams of data to train them to make decisions and carry out actions on their own, without human intervention. But AI experts have said that such decisions are often only as good as the data used to train the models.

Singapore is investing in frameworks and platforms to help different parties exchange data in a secure manner while ensuring data privacy, said Mrs Teo.

This includes the Singapore Financial Data Exchange (SGFinDex) launched in 2020. It pools users' financial data from across the public and private sectors on a single platform.

Mrs Teo said aggregating data in this way "opens up new possibilities for banks and other financial service providers". For instance, they can offer new AI-enabled services, such as predictive analytics, to help customers manage their finances.

Singapore is also working with international partners on supporting data flow across borders to fuel AI innovations, she said.

But there are instances in which data sharing is not possible due to regulatory, ethical and commercial sensitivities.

To address this, Singapore is investing in techniques like "federated learning", said Mrs Teo. Such learning lets different parties train AI models together without having to exchange raw data.

AI Singapore, the national AI research and development programme, has developed one open-source platform for this called Synergos, which Mrs Teo said makes federated learning more user-friendly and accessible.

Join ST's Telegram channel and get the latest breaking news delivered to you.