AI in classrooms and skills students still need to practise, Snowflake's data layer, and Apple Pay beside the national rail in the Philippines

This Week's Analysis

As AI enters APAC classrooms, what skills should students still practise?

Short Reads

1. Snowflake in ASEAN: The data layer beneath the AI boom

   Southeast Asia's AI ambitions keep running into a very unglamorous problem: the data is scattered everywhere. Snowflake is selling a way to bring that data together, while competing with hyperscalers that increasingly offer similar services on their own clouds. This Short Read looks at what Snowflake is actually selling in ASEAN, and who else is selling it.

   Read on Relay Asia →
2. Apple Pay landed in the Philippines on cards. The national payment rail was already there.

   Apple Pay has arrived in the Philippines on Visa and Mastercard cards. The pitch is convenience. But Filipino consumers already have a way to pay without cash: QR Ph lets them scan a code through participating banks and e-wallets. So Apple Pay is arriving as another wallet, rather than filling a gap in the country's payment infrastructure. This Short Read looks at what happens when a global wallet arrives alongside a national payment rail.

   Read on Relay Asia →

Suggested Reads

• Bill Gates says governments are way behind on AI. Bill Gates told Reuters that “governments are way behind” on AI, and that no government is “nearly as deep on this as they have to be.” That's not exactly a new argument. The more interesting part is where he thinks the tools need to reach. His foundation says it will spend at least USD 1 billion over two years on areas including education, healthcare and agriculture, as well as the language and data foundations that AI models depend on.
• Grab is buying a controlling stake in Atome Financial. Grab is buying 60% of Atome Financial for USD 1.49 billion. Atome already lends across Singapore, Malaysia, the Philippines, Indonesia and Thailand. Grab plans to buy the remaining 40% roughly two years after the deal closes, at a price that has yet to be determined. The deal is another interesting look at how a superapp moves deeper into consumer finance.
• Thailand is putting an AI sales assistant in front of a bank loan. Nok Krasip, or “Whispering Bird”, started as an AI sales assistant inside Krungthai Bank's Tung Ngern app. Now it is being put in front of a loan. Traders can submit three months of sales data for a credit assessment without visiting a branch. Krungthai is offering up to THB 500,000 for small traders and up to THB 3 million for small and medium-sized businesses. This is the kind of AI deployment I find more interesting than another chatbot demo.
• Google Cloud opened a Singapore engineering centre, not another support office. Google Cloud has opened its Singapore Engineering Center, its first in Southeast Asia, alongside the DeepMind lab. What stands out is the ambition. This is positioned as a product engineering centre, with work built in Singapore for use elsewhere. Early collaborations include multilingual models with Grab and agentic workflows with DBS. Singapore is increasingly being used not just as a market for AI products, but as a place to build them.
• AEON Malaysia is profiling customers by life stage, not just their last basket. AEON360, the data arm of AEON Co. (M) and AEON Credit, is trying to understand customers by where they are in life, rather than simply what they bought last. Their example is a parent: baby formula, a personal loan and a confinement-centre booking, all read together. It's a good illustration of what becomes possible when retail, mall and financial data start being viewed as one customer profile.
• A short sketch about maxing out the monthly AI tokens by the 6th. And finally, something a little less serious: this sketch imagines what happens when you've blown through your monthly AI token allowance by the 6th and need to borrow enough from a colleague for just one more question. The question, naturally, is a Millennium Prize problem.