Recent Posts
15 posts
The Low-Altitude Economy — and the Model Malaysia Needs to Fly It
Drones spraying plantations and air taxis crossing skylines run on the same stack: an aircraft, an airspace management system, and a regulator willing to say yes to routine, unsupervised flight. The low-altitude economy is what that stack looks like once it starts generating revenue. This is a study of how China, the US, the EU, and Singapore are each running that system, what R&D work actually carries it forward, and what Malaysia needs to build to turn its existing pilots into a national industry.
Large Language Models, Agentic Computing, and the Builders Malaysia Still Needs
How next-token prediction became agentic computing — tracing the mechanism from LLMs to tool-using agents, benchmarking GPT-6 Astra against Claude Fable's 2026 releases, contrasting America's enterprise-gated approach with China's open-weight diffusion strategy, and arguing Malaysia's ILMU/Ilmu-Nemo-30B agentic models are the right sovereignty bet — not a chatbot flex.
What the Hydrogen Inhaler Costs the Rakyat — and Why Local Universities Keep Losing This Race
In the third week of August 2026, the Ministry of Health warned Malaysians about a device sold with no registration and no evidence, aimed at people managing chronic illness. It sold anyway. The question worth sitting with isn't just "how was this allowed to happen" — it's "why does the informal market keep winning the race to reach ordinary Malaysians before our own universities do."
The Real Journey of AI Commercialisation — and the Model Malaysia Needs to Run It
Most AI research never becomes a product. It becomes a paper, a patent filing, or a conference poster, and then it stops. The organisations and countries that have figured out how to consistently push discoveries past that stopping point are not doing something mysterious — they are running a repeatable system that carries an idea through validation, partnership, capital, and market entry. This is a study of that system: how it works, who runs it well, and what Malaysia would need to build to run it too.
Deep Tech and the Commercialisation Models That Actually Work
Most innovation gets built on top of what already works — a faster app, a cheaper feature, a better interface. Deep tech starts somewhere else entirely: in a lab, with a scientific or engineering question that has no guaranteed answer, let alone a guaranteed market. The distance between that question and a paying customer is where most deep tech dies — not from bad science, but from science that never learned to speak the language of an industry problem.
Crossing the Valley: Why Great University Research Dies Before It Reaches the Market
Every research university produces more good ideas than it will ever turn into products. This is not a scandal — it is close to the normal condition of applied science. What separates innovation-rich economies from innovation-poor ones is not how many discoveries a lab makes, but how many of them survive the unglamorous middle distance between a working prototype and a paying customer. That distance has a name, a shape, and — increasingly — a set of institutions built specifically to shorten it.
No posts match your search