From Networking to the Neurodigital Ecosystem

Technology Evolves Because We Understand Ourselves

Every major technological revolution begins with a deeper understanding of what it means to be human. Technology has never advanced simply because processors became faster or software grew more sophisticated. Those developments matter, but they are rarely the true starting point. Innovation begins when we discover something new about ourselves.

Every human capability that we learn to observe, understand, and formalize becomes, at least in part, reproducible through technology. This has happened repeatedly throughout history: first with physical labour, then with mathematical computation, later with industrial automation, and today with increasingly sophisticated cognitive functions. Each time we learn to delegate part of a human capability, we also create the conditions to redirect our attention toward exploration, creativity, cooperation, and the construction of meaning. Technology does not evolve by replacing human intelligence; it evolves by expanding the space in which human intelligence can continue to grow. Artificial intelligence, in this sense, should be understood as the latest chapter in a much longer historical process.

Public attention naturally gravitates toward what appears new. Today that attention is focused on increasingly powerful AI models, reinforcing the belief that the future of digital transformation depends primarily on computational scale. The history of computing, however, suggests a different pattern. Innovations that initially dominate the conversation rarely remain at its centre. As they mature, they gradually become embedded capabilities of the infrastructure itself. Operating systems, databases, networking, Internet technologies, virtualization, cloud computing, and, more recently, cybersecurity have all followed this trajectory. None of these innovations became less important; on the contrary, they became so fundamental that they disappeared into the infrastructure, providing the foundation upon which subsequent generations of technology could emerge.

I believe artificial intelligence will follow the same path. Today we experience AI as a product; tomorrow we are likely to experience it as an intrinsic capability of digital infrastructure. This may seem like a subtle distinction, yet it fundamentally changes the way we should think about the future.

For this reason, I find myself paying less attention to the competition between AI models and more attention to the evolution of digital infrastructure. Models will undoubtedly continue to improve, but the most significant transformation may occur elsewhere. Digital infrastructures are no longer passive systems designed simply to transport information. They are evolving into environments capable of observing events, coordinating distributed processes, integrating knowledge generated across different contexts, and supporting decisions precisely where those decisions need to be made.

If this trajectory continues, the next technological paradigm may be defined less by computational power than by the ability of digital infrastructures to distribute intelligence throughout the entire ecosystem, making it more contextual, less centralized, and increasingly embedded in everyday interactions.

Looking at Infrastructure Instead

This is why I find myself looking less at AI models and more at digital infrastructure.

Throughout the history of computing, infrastructures have almost always anticipated applications. They create the conditions that allow new technologies to emerge, support their adoption, and eventually become so deeply embedded that they disappear from everyday attention. By the time a technology becomes invisible, it has usually become indispensable. The real transformation, in other words, often begins long before it becomes visible to the market.

From this perspective, some companies are interesting not because they predict the future, but because they make an ongoing transformation easier to observe. Cisco is one of them. Its relevance has little to do with competing to build the most powerful AI model. What makes the company particularly interesting is the direction of its evolution. Over the past decade, networking, observability, digital identity, cybersecurity, the Internet of Things, data platforms, and artificial intelligence have gradually stopped evolving as separate domains. They are converging into a single digital architecture, suggesting that the future of AI may depend less on individual models than on the environments in which those models operate.

Cisco's acquisitions illustrate this transformation remarkably well. Jasper extended its reach into the Internet of Things; ThousandEyes and AppDynamics expanded visibility across networks and applications; Duo Security strengthened digital identity through a Zero Trust approach; Splunk brought telemetry, cybersecurity, and data analytics into the same architectural framework. Considered individually, these acquisitions resemble the natural expansion of a technology portfolio. Viewed together, however, they reveal a much broader strategy: building an infrastructure capable not only of connecting devices, but of sensing what is happening across the entire digital environment, transforming telemetry into contextual intelligence, and supporting decisions where events actually occur.

This distinction matters.

For decades, digital infrastructures were designed primarily to move information from the edge toward a centralized point of control. Increasingly, they are being asked to do something different: observe locally, respond locally, and allow those local decisions to contribute continuously to the intelligence of the entire system. The infrastructure is no longer simply transporting information; it is becoming an active participant in how information is interpreted, coordinated, and transformed into action.

If this transition continues, the greatest opportunities may emerge not from the infrastructure itself, but from the ecosystem that will grow around it. As artificial intelligence becomes an embedded capability rather than a standalone product, new generations of applications, specialized platforms, and integration partners will be required to translate distributed intelligence into operational, organizational, and strategic value. The role of systems integrators will evolve accordingly. Their task will no longer be limited to connecting heterogeneous technologies, but to designing environments in which networking, identity, security, observability, IoT, data platforms, and artificial intelligence function as parts of a single adaptive system.

The future may belong less to those who build individual AI models than to those who learn how to orchestrate intelligence across an entire ecosystem.

Beyond Cognitive AI

If digital infrastructures are indeed evolving in this direction, the next challenge may no longer be technological alone. It may also require a broader understanding of intelligence itself.

For decades, digital technologies have drawn inspiration primarily from cognitive psychology. Artificial intelligence has learned to recognize patterns, classify information, predict outcomes, and support individual decision-making with extraordinary effectiveness. These capabilities have transformed entire industries and will continue to do so. Yet the infrastructures now emerging appear to demand something more. They are no longer expected to support a single centralized decision, but thousands of contextual decisions made simultaneously across factories, hospitals, cities, supply chains, industrial systems, and connected devices. Their challenge is not simply to calculate more efficiently, but to coordinate decisions that arise in different places while preserving the coherence of the whole.

This is where relational psychology and contemporary neuroscience offer a different perspective. Rather than describing intelligence as the activity of an isolated mind processing information, they suggest that intelligence emerges through the continuous interaction between perception, experience, relationships, and context. Decisions are rarely the product of information alone. They are shaped by previous experience, influenced by relationships, constantly revised through interaction with the environment, and ultimately transformed into shared understanding.

If this is how human intelligence develops, future digital infrastructures may eventually be inspired by the same principle. Their role will extend far beyond processing data or generating increasingly accurate inferences. They will need to enable networks of distributed intelligence capable of observing their local environment, responding where events occur, and continuously contributing those local decisions to the understanding of the system as a whole. The challenge, in other words, is shifting from computational performance to collaborative intelligence.

Seen from this perspective, even the word infrastructure begins to feel inadequate. What is emerging resembles less a traditional IT architecture than a living nervous system. Networks transmit signals. Identity establishes trust. Observability allows the system to perceive itself. IoT senses the surrounding environment. Artificial intelligence interprets those signals. Security preserves the integrity of the whole. Data platforms integrate information generated across thousands of distributed contexts. Individually, each technology solves a specific problem; together, they begin to behave like an adaptive organism.

The analogy with the human nervous system is more than a metaphor. Intelligence does not reside in a single centre that controls everything. It emerges from the continuous dialogue between the central and peripheral nervous systems, where millions of local interactions perceive the environment, respond in real time, and contribute to the stability of the organism as a whole. Digital infrastructures appear to be evolving in much the same way. Their value will depend less on the intelligence of any single model than on their ability to orchestrate relationships among people, systems, and distributed forms of artificial intelligence.

Perhaps this is the beginning of what we might call a neurodigital ecosystem.

The Rise of Proximity Intelligence

If this evolution continues, the most significant transformation will not be the emergence of ever more powerful AI models or increasingly larger data centres. It will be the way intelligence itself becomes distributed across the digital ecosystem.

For decades, digital infrastructures followed a relatively simple logic. Information was collected at the edge, transferred to a central system, and transformed there into decisions. The architectures now emerging suggest a different model. Intelligence is gradually moving closer to where events occur. Factories, hospitals, cities, energy grids, transportation systems, and connected devices will increasingly observe their own environments, interpret local conditions, and respond in real time while remaining connected to broader organizational objectives.

This evolution does not eliminate the centre; it redefines its role. Much like the human nervous system, where the brain does not control every individual action but continuously integrates information generated by millions of peripheral interactions, future digital infrastructures may become ecosystems in which intelligence emerges from the dialogue between local autonomy and global coordination. The value of the system will depend less on the power of a single model than on the quality of the relationships connecting distributed sources of intelligence.

This is how I imagine the emergence of proximity intelligence. It is not an alternative to centralized intelligence but its natural evolution. Local intelligence enriches the whole by making it more responsive, more contextual, and ultimately more adaptive.

If this perspective proves correct, value creation will shift accordingly. Competitive advantage will arise not only from infrastructure itself, nor exclusively from access to increasingly capable AI models, but from the ability to integrate heterogeneous technologies into environments that learn, adapt, and coordinate decisions across complex organizations. Systems integrators will therefore assume a profoundly different role. Rather than simply connecting technologies, they will increasingly design collaborative ecosystems in which networking, digital identity, cybersecurity, observability, IoT, distributed data platforms, and artificial intelligence operate as parts of a single adaptive system. Their expertise will extend well beyond engineering into organizational architecture, because future infrastructures will increasingly mirror the way organizations themselves learn, collaborate, and evolve.

Perhaps this brings us back to where every technological revolution begins.

We first discover something new about ourselves.

We then learn how to formalize part of that discovery.

Eventually, we delegate part of it to technology.

Every delegation expands rather than diminishes the possibilities of human intelligence, allowing us to redirect our efforts toward new forms of exploration, creativity, cooperation, and meaning making.

Artificial intelligence may simply represent the next chapter in that long history.

Its greatest contribution will not be replacing human intelligence.

It will be creating the conditions for human intelligence to continue evolving.

Because technological progress has never truly been about replacing human beings. It has always been about expanding our capacity to explore, create, cooperate, and construct meaning.

If the next generation of digital infrastructures is indeed moving toward distributed intelligence, one concept will become increasingly central to understanding this transformation: inference.

The word appears constantly in discussions about artificial intelligence, yet its meaning changes significantly depending on whether we approach it from logic, computer science, neuroscience, or psychology. Understanding these different perspectives helps clarify not only how AI works, but also what will continue to distinguish human intelligence in the years ahead.

Appendix – What Do We Mean by Inference?

Few words have become as central to today's discussion of artificial intelligence as inference. Yet the same word carries remarkably different meanings across disciplines. Rather than competing definitions, these represent complementary ways of understanding how information becomes knowledge and, ultimately, how knowledge becomes action.

In logic, inference is the process of deriving a conclusion from one or more premises through deductive or inductive reasoning. The emphasis lies on the validity of the reasoning itself: does the conclusion legitimately follow from what is already known?

In computer science and machine learning, inference refers to the operational phase of a trained model. Once learning has taken place, the model applies what it has learned to new data in order to generate a prediction, a classification, or a decision. The objective is not to establish truth, but to produce the most reliable response possible within a given context.

In neuroscience, the concept takes on a broader meaning. According to theories such as Predictive Processing and Active Inference, the brain continuously generates hypotheses about the world, compares them with incoming sensory information, and updates them whenever experience contradicts its expectations. Inference therefore becomes an ongoing adaptive process that enables living systems to navigate uncertainty.

Relational psychology introduces a further dimension. Its interest lies not only in how people generate inferences, but in how those inferences evolve through relationships. Human beings routinely make decisions with incomplete information, tolerate ambiguity, revise their assumptions, suspend judgment, negotiate meaning with others, and sometimes abandon previous interpretations altogether in order to explore entirely new possibilities. In this perspective, intelligence is not simply the production of inferences; it is the continuous construction of shared meaning.

Taken together, these perspectives reveal a remarkable progression: from logical validity to computational prediction, to biological adaptation, and finally to relational meaning-making.

This progression also offers a useful lens through which to understand the future of artificial intelligence. AI will become increasingly effective at generating inferences. Human intelligence, however, will continue to distinguish itself by integrating those inferences, placing them within a broader context, questioning them when necessary, and transforming them into understanding, cooperation, and innovation.

If the next technological paradigm is indeed one of distributed intelligence, the central challenge will no longer be producing more accurate inferences. It will be creating the conditions in which inferences generated by people, intelligent systems, and different contexts can interact, challenge one another, and gradually converge toward shared understanding.

Inference may become the language of artificial intelligence.

Meaning making will remain the language of human intelligence.

The future will depend on our ability to transform distributed inferences into shared meaning.