From centralised AI to distributed intelligence: implications for bioinformatics

This article examines a possible transition in artificial intelligence from centralised, cloud-based systems toward smaller, specialized models operating locally on personal devices, laboratory instruments, medical equipment, and industrial machinery. Drawing on the historical shift from mainframe computers to personal computing, it argues that local AI may complement, and in some applications partly replace large centralized models by providing lower latency, greater privacy, reduced operating costs, and less dependence on cloud connectivity. Although hyperscale data centers will remain essential for training frontier models and addressing complex problems, the growth of local.AI could challenge current infrastructure-focused business models and lead to market consolidation and repricing. The article further explores the implications of this transition for bioinformatics and the life sciences. It argues that bioinformatics should become an integral part of life-science education and research, enabling scientists not only to use AI tools but also to understand, validate, and develop them. Universities and research institutions should therefore support interdisciplinary training and research environments in which specialised AI solutions can be designed and deployed for biotechnology, medicine, agriculture, food production, and environmental monitoring. The article concludes that bioinformatics can serve as a vital bridge between AI and biology, allowing the life sciences to actively shape the transition from centralised, general-purpose AI to distributed, specialised local intelligence.