The conversation in Singapore this week started with a basic fact: the technology works. But it doesn’t work in a vacuum. That much became clear as healthcare IT professionals and policymakers sat down to talk about artificial intelligence in Asia.
The meeting on June 28 wasn’t about algorithms or processing power. It was about what happens before the AI ever gets turned on.
The first step, as the discussion laid out, is data. Clean data. Standardised data.
Without it, algorithms learn wrong. They pick up biases.
They make mistakes. For a family in Manila or a clinic in rural Thailand, that means a misdiagnosis. A missed screening.
A wrong treatment path. The experts were blunt: you cannot skip this step. One systems architect at the roundtable put it simply.
“You can’t build a skyscraper on sand.” Next came the question of infrastructure. Digital records in many Asian hospitals are incomplete.
Some are paper-based. Some exist in different formats that don’t talk to each other. The roundtable heard that a hospital cannot leapfrog straight to AI diagnostics if its digital backbone is weak.
The technology, no matter how advanced, becomes useless if it doesn’t fit into existing workflows. Staff need training.
Systems need integration. These are not glamorous problems. They are foundational ones.
The timeline then moved to the systemic layer. Deploying AI in isolated pockets was dismissed as a dead end. Instead, the call was for integrated strategies that connect hospitals, clinics, and community health workers.
Interoperability became the keyword. A patient’s data must travel with them, securely, from one provider to the next.
A rural mother should not have fewer options than someone in a city. That is what equity looks like in practice, the speakers argued. The discussion didn’t stop at technical fixes.
It circled back to people. The promise of AI in healthcare has felt distant to many communities across Asia.
The experts said the real challenge is not the technology itself. It is getting the foundations right. That means practical, people-first steps.
It means training staff. It means standardising records. It means building trust.
The roundtable did not offer a neat conclusion. There was no single solution.
Instead, there was a clear sequence of priorities. First, fix the data. Second, build the infrastructure.
Third, connect the systems. Fourth, train the people.
Only then does AI have a chance to reach the families and patients who need it most. For a hospital in Manila or a clinic in rural Thailand, the path forward is not about skipping ahead. It is about doing the hard, unglamorous work of laying a solid foundation.
The skyscraper can wait. The sand needs to be cleared first.
