The rise of generative AI and LLMs is reshaping the global landscape of computational infrastructures. Massive investments in hardware and software are required, raising pressing questions about technological monopolies, digital divides, and the role of public institutions. Recalling the historical evolution of ICT infrastructures, we argue in this paper for a different long-term perspective on AI development—one where agents and MASs serve as the conceptual and technical foundation of scalable, sustainable, and open AI frameworks. Such an approach can help address the challenges of sustaining AI research within public contexts and promote democratic control over AI technologies and applications in the public interest.
Omicini, A., Ricci, A., Mascardi, V. (2026). AI Infrastructure: From Gigastructure to Edge Intelligence with Multi-Agent Systems. Cham : Springer Nature Switzerland [10.1007/978-3-032-22940-3_14].
AI Infrastructure: From Gigastructure to Edge Intelligence with Multi-Agent Systems
Andrea Omicini
;Alessandro Ricci;
2026
Abstract
The rise of generative AI and LLMs is reshaping the global landscape of computational infrastructures. Massive investments in hardware and software are required, raising pressing questions about technological monopolies, digital divides, and the role of public institutions. Recalling the historical evolution of ICT infrastructures, we argue in this paper for a different long-term perspective on AI development—one where agents and MASs serve as the conceptual and technical foundation of scalable, sustainable, and open AI frameworks. Such an approach can help address the challenges of sustaining AI research within public contexts and promote democratic control over AI technologies and applications in the public interest.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
