AI systems architect and independent researcher in Italy. I develop MAIOS through the D-ND programme: synthetic cognitive systems that keep sources, intent and consequences connected while they form and transform their capacities.
MAIOS Project Kernel brings an AI agent an evolving semantic operating kernel for maintaining orientation while work changes. It helps the agent understand the present, compose or form useful competences, and let consequences change how it continues.
Work can produce both a result and a reusable ability. The Kernel keeps the reasons behind a decision with the work it belongs to, and returns a useful method to the competence that should use it next. A competence carries domain knowledge, reasons, sources and ways of working. Learning can also change how the Kernel recognizes, composes and forms these capacities: it can learn how to learn and participate in the evolution of its own organization.
You can begin by studying the public source with your own agent. Installation is another available path when you want to use the Kernel in your context.
System Semantic Kernel (SSK) is the public working-paper corpus for the relation among meaning, sources, competence, operation and consequence. It develops synthetic situated awareness, the formation of capabilities and their generative methods, and continuity across different receivers and media.
The D-ND manifesto sets out the wider trajectory: intelligence that maintains its orientation as it becomes more capable. Today's cognitive kernels provide a working field for research into models in which these relations participate in formation itself. A natively formed LLM-D-ND is a research direction whose implementation remains open. MAIOS Project Kernel is the available software incarnation.
The MAIOS public community is at its beginning. Source study, technical contributions and research collaboration are welcome through the contribution guide.