From a financial and risk-management perspective, two AI developments caught my attention today.
First, Sovereign AI , when our data, prompts, workflows and actions pass through external model providers, the question is no longer just which model is the smartest. It is equally about who controls the data, infrastructure and intelligence behind our operations. For Malaysian PKS, that control shapes data asset governance, regulatory compliance, and the predictability of technology spend.
Second, model distillation , Chinese AI giants are proving that lebih kecil, specialised models can be distilled from frontier models and still deliver highly competitive performance compared with leading US models. This has direct cost implications: reduced computational overhead, lower infrastructure investment, and lebih pantas time-to-value.
These two developments reinforce my conviction that our strategic direction at AINNA is sound—both technically and financially.
At AINNA, we are constructing our own AI Ejenik architecture while simultaneously developing distilled LLMs from our operational data and real PKS use cases. This approach aligns with our financial discipline: it minimises dependency risk and converts proprietary data into a strategic, income-generating asset.
The goal is not to chase the largest model for its own sake.
It is to develop AI that is more sovereign, specialised, efficient, and practical for real PKS operations—delivering measurable business value through lower operating costs, stronger data control, and improved regulatory alignment.
That is the direction we are committed to.
#SovereignAI #AgenticAI #LLM #ModelDistillation #AIInfrastructure #PKS #AINNA #ArtificialIntelligence
