I was delighted to participate in the first Boston TECH WEEK by a16z. With hundreds of events taking place across Boston and Cambridge, the week created a distributed space for founders, investors, policymakers, researchers, operators and innovators to engage with the current frontier of technology.

​

AI was, unsurprisingly, central to many conversations. But what made the week particularly relevant was not only the technology itself. It was the way discussions moved across sectors and disciplines, from manufacturing, energy and enterprise systems to healthcare, biotech, investment, governance and geopolitics, and kept returning to a deeper questions: What does it take for organisations, sectors and ecosystems to make effective use of AI as it becomes embedded in real systems? and, even deeper, how is AI changing the systems in which organisations operate?

​

For me, this was not an abstract question. It connects directly with my work across technology implementation, digital transformation, innovation policy, health technologies and international innovation ecosystems. Especially, with a whole career around supporting organisations in digital transformation.

AI is not simply another tool being added to existing organisations. Leaders, in fact, need to realise that it is not just a tool, it is becoming a real workforce within their organisations, reshaping how work is organised, how decisions are made, how knowledge is captured, how systems are governed, and how value is created across sectors.

​

That is why the next phase of AI will not be defined only by technical progress. It will be defined by readiness, judgement and implementation capacity.

​

The confidence paradox of AI implementation

​​

Of course AI is already being implemented. That is not the question.

The harder question is what kind of implementation we are seeing and whether organisations understand what AI is changing.

​

Many companies are testing tools, launching pilots, embedding AI into workflows, or experimenting with agentic systems. But introducing AI into an organisation is not the same as understanding how it changes capability, accountability, leadership, knowledge management and operating models.

​

This is where the confidence paradox becomes relevant. At the beginning, many organisations feel they understand AI because they can see use cases, tools and immediate productivity gains. Then implementation begins, and they realise how much they do not yet understand: their data, workflows, decision rights, governance structures, operational costs, institutional knowledge and risk exposure. Over time, the organisations that progress are those that build a more mature confidence, one grounded in tested capability rather than enthusiasm.

​

This is not entirely new. Digital transformation, automation, data platforms and IoT raised similar questions. But AI increases the speed, scope and intensity of the challenge because it touches not only processes, but judgement, knowledge, coordination and decision-making. The risk, many organisations do not know how those decisions are taken.

​

Recent reports from MIT, Cisco and Gartner point in this direction: many organisations are investing and experimenting, but far fewer are fully prepared to capture value at scale.

​

For me, this is where the conversation becomes more important. Not “is AI being adopted?”, clearly, it is. But what is AI changing, and are organisations, sectors and societies prepared for that change?

​

Readiness is not only technological

​

For more than a decade, my policy and advisory work has focused on the enabling conditions for technology adoption: how societies, institutions and sectors build the policy and incentives, governance, capabilities, infrastructure, leadership and trust required to benefit from innovation.

​

This question has become even more urgent with AI. The issue is not only whether emerging technologies are ready to be deployed, but whether the surrounding systems are prepared to absorb, implement and govern them responsibly.

​

This also connects with my experience integrating and implementing technologies in complex, often highly regulated environments. The challenge is rarely the technology by itself. The more difficult question is whether organisations can create the conditions for adoption, use and long-term value.

​

This was also evident through Camnexus IoT's own experience. When we developed our AI-enabled IoT platform in 2019, working at the intersection of emerging technologies, data, infrastructure and operational systems, it became clear that the wider context mattered as much as the technology. And that was true for both developed countries as well as emerging regions.

​

AI-native thinking is powerful, but many organisations expected to work with AI are not AI-native. Some are not even technology-driven at their core. They are legacy companies, including public institutions, health systems, utilities, manufacturers, infrastructure operators and regulated organisations with existing systems, constraints, incentives and knowledge structures.

​

The real challenge is therefore not only how to build AI. It is how to integrate AI into organisations and systems that were not designed around AI. And in many cases, have not yet gone through a full digital transformation journey. This is where expectations of “leapfrogging” need to be treated carefully.

​

Resilience, not only efficiency

​

One theme that stood out strongly was the shift from efficiency to resilience, particularly in industrial and infrastructure-related sectors - aligned with the global scenarios, which I observed also at the The World Bank Group / International Monetary Fund Spring Meetings that I attend in April this year.

​

In manufacturing, energy, utilities and infrastructure, AI is often discussed in relation to productivity, optimisation, predictive maintenance or cost reduction. Those are important. But the more strategic question is how technology can help systems become more adaptive, distributed and resilient.

This connects directly with the vision behind Camnexus. We have approached technology as a platform for Sustainable Development: a way to close innovation gaps and support organisations in addressing complex challenges linked to infrastructure, energy, industrial transformation, environmental pressures and long-term resilience.

​

For us, this is not a new agenda. Long before AI became the dominant conversation, the core issue was already how to use emerging technologies to strengthen systems that matter, not only to make them more efficient, but to make them more adaptive, inclusive and resilient.

​

This is especially important in industrial environments. These systems operate within complex constraints: legacy equipment, fragmented data environments, safety requirements, procurement cycles, regulatory obligations, workforce transitions and physical infrastructure. In many cases, critical knowledge is not fully codified. It sits with the people who understand how systems behave under stress.

​

Another important message was the shift from AI as a tool to AI as part of the workforce. As enterprise applications increasingly embed agents, organisations will not only use AI; they will need to coordinate, supervise and lead AI-enabled workflows. That changes the adoption question. It is no longer only about whether a model performs well, but whether people, processes and systems are ready to work with agentic capabilities in practice. New processes will emerge, new models and organisations.

​

That is where decentralised systems and collaborative intelligence become important. In complex industrial settings, intelligence is rarely located in one place. It is distributed across machines, data systems, operators, engineers, suppliers, infrastructure and decision-makers. AI can create value when it strengthens coordination across these distributed sources of knowledge: improving visibility, supporting decisions closer to where operational realities sit, and helping systems become more adaptive and resilient.

​

In these contexts, value does not come only from algorithmic performance. It comes from connecting AI with operational knowledge, system constraints, decentralised decision-making and the realities of resilience.

​

In healthcare and biotech, translation is the real test

​

Healthcare, pharma and biotech make the AI conversation more concrete because they expose a tension that is easy to underestimate: AI moves fast, but health technologies move through evidence, regulation, trust, production capacity and adoption pathways that cannot simply be compressed.

​

AI may accelerate discovery, diagnostics, regulatory intelligence, clinical operations and supply-chain decisions. But in health, speed alone is not progress.

​

A promising technology still has to be validated, financed, regulated, manufactured, trusted and adopted within systems that are complex, highly regulated and uneven in their capacity.

The same applies in biotech. Scientific promise is only one part of the story. Strategic investment can move technologies forward, but capital alone does not solve translation. What matters is whether the right partnerships, regulatory pathways, production capacity, health-system needs and market access conditions are aligned at the right time.

​

That is where AI, biotech and health systems meet the same reality: value is not created when a technology looks promising, but when the system around it can turn that promise into use.

​

For health technologies, that means better decisions, stronger institutions, improved access and real health-system value, not simply more advanced tools.

​

Governance and policies: principles to traceability

​

For several years, much of the discussion has focused on principles: responsible AI, fairness, transparency, accountability, privacy and human oversight. These principles remain important, but they are no longer sufficient on their own.

​

AI systems as they become embedded in operations, supply chains, finance, healthcare, public services and enterprise decision-making, organisations need to move from policy statements to traceable execution.

​

That means being able to understand what an AI-enabled system did, what context informed the decision, who is accountable, how the decision can be inspected, and what happens when the system produces an unexpected outcome. This becomes even more important when organisations rely on multiple vendors, platforms and third-party operators across the AI value chain.

​

This matters in every sector, but especially in complex and regulated environments.

In healthcare, AI-enabled decisions may affect patient pathways, clinical operations, resource allocation, access to technologies and health system performance. In manufacturing and energy, they may affect operational continuity, safety, maintenance, supply chains and infrastructure resilience. In public-sector settings, they may influence service delivery, eligibility, prioritisation and institutional trust.

 

The more AI enters consequential decision-making, the more governance needs to become operational.

 

Traceability is therefore becoming central to AI readiness. Without it, organisations may struggle to defend decisions, manage risk, build trust or scale AI responsibly.

​

Technology strategy and geopolitical reality

​

Another important layer is geopolitics. Technology decisions cannot be separated from global political and economic dynamics.

​

AI, digital infrastructure, data flows, supply chains, semiconductors, cybersecurity, regulation, industrial policy and investment are increasingly connected. For technology executives and organisations working internationally, these are not abstract geopolitical issues. They shape product strategy, market access, operational risk, data governance, procurement choices and long-term growth.

​

This is especially relevant for organisations building innovation capacity across countries and sectors. The design of digital transformation strategies, AI pathways and technology partnerships must now account for cross-border risk, regulatory divergence, security expectations, supply-chain exposure and institutional trust.

​

As one of the top innovation ecosystems, Boston/Cambridge brings together science, venture, universities, enterprise technology, healthcare, deep tech and policy thinking in a concentrated ecosystem.

​

Building on the success and learnings from top innovation ecosystems, like Cambridge UK, my own work through Camnexus focused precisely on supporting innovation ecosystem development and strengthening across more than 24 countries. This has involved working with policymakers, universities, entrepreneurs, investors, companies and international organisations to help identify strengths, close capability gaps and develop the enabling conditions for innovation to be adopted and sustained.

Independent of the specific technology or innovation, the underlying question is often the same: are organisations, sectors and societies prepared for it?

​

That question is becoming particularly relevant now, given the speed of AI development and the pace at which new tools are moving into markets, institutions and everyday decision-making. The issue is not only whether the technology is ready. It is whether the surrounding systems are ready to absorb, implement and govern it responsibly.

​

From technology readiness to system readiness

 

One of my main reflections from Boston Tech Week is that AI readiness cannot be understood only at the level of the individual organisation.

​

It also needs to be understood at the level of systems and ecosystems.

​

Companies may need data infrastructure, internal skills and governance mechanisms. But broader environments also need enabling policies, investment channels, technical talent, applied research, procurement pathways, standards, trusted intermediaries, sector-specific knowledge and communities that connect innovators with adopters.

​

This is particularly important when working across sectors such as healthcare, biotech, manufacturing, energy, utilities, infrastructure and public services. In these environments, technology adoption depends on alignment between multiple actors. No single company or institution can solve the readiness problem alone.

​

That is why innovation ecosystems matter.

​

They determine whether technologies remain isolated pilots or become part of wider transformation. They shape whether early adopters can influence markets. They affect whether companies can access the right partners, whether policymakers understand implementation constraints, whether investors recognise adoption barriers, and whether institutions have the capacity to use new tools effectively.

​

The AI moment makes this work even more relevant. As AI accelerates, the gap between technological possibility and institutional capacity becomes more visible.

​

The next frontier is implementation capacity

​

The first phase of the current AI cycle has been dominated by experimentation, excitement and rapid tool development. That phase is still ongoing, but it is no longer enough.

​

The next phase will depend on implementation capacity.

 

In manufacturing and energy, this means resilience, collaborative intelligence and operational knowledge. In enterprise AI, it means defensibility and durable value. In healthcare and biotech, it means translation across science, regulation, investment and health-system needs. Across all sectors, it requires traceability and accountability. In geopolitics, it means understanding the strategic conditions under which technologies are developed, adopted and scaled.

​

The opportunity is significant, but so is the implementation challenge.

​

AI will not create lasting value simply because it is powerful. It will create value where organisations and ecosystems are ready to absorb it, govern it and use it to strengthen real capabilities.

​

That is the frontier that matters now.

​

Boston Tech Week was a valuable space to engage with these questions across sectors and perspectives. It was energising to connect with founders, investors, policymakers, operators, researchers, security experts and enterprise leaders working through these challenges from different angles.

​

The human factor as connective fabric

​

One theme I have emphasised for many years, including in my TEDx talk in 2018, is that innovation is not only about technology. It is about people.

​

Boston Tech Week brought that point back in a very current context. Innovation ecosystems are built around movement: the movement of knowledge, people, ideas, capital, capabilities and opportunities.

​

Beyond the formal sessions, the conversations and networking spaces made this visible in a very practical way.

​

Many discussions were not only about technologies, markets or investment. They were also about people navigating change: founders building new ventures, professionals repositioning themselves, companies looking for direction, and individuals trying to find new communities, purpose and identity in a rapidly shifting technology landscape.

​

This is particularly relevant now because the same technologies creating new opportunities are also forcing people and organisations to rethink their place in the ecosystem. AI and digital transformation are creating new roles and ventures, but they are also contributing to disruption, reinvention and the need for new forms of connection.

​

That human dimension is not separate from innovation ecosystems: it is central. It is part of the connective fabric that allows ecosystems to absorb change, circulate knowledge, build trust and create new opportunities.

​

Leaving Boston and heading back to Washington, DC, I am convinced that the future of AI will depend not only on technological progress, but on the ability of organisations and ecosystems to translate that progress into resilient, trusted and useful capabilities.

​

#BOSTechWeek #AI #DigitalTransformation #InnovationEcosystems #EnterpriseAI #AIGovernance #Manufacturing #Biotech #HealthInnovation #DeepTech #sustainabledevelopment #globalchallenges #climatetech #Inclusiveinnovation