Following the publication of our recent World Economic Forum article, I am pleased to share here below the authors' draft, developed in collaboration with @Dr. Giovanna Di Mauro.

Culture and the production of meaning

In Crime and Punishment, Fyodor Dostoyevsky places moral responsibility at the centre of human freedom, suggesting that knowledge alone is insufficient without ethical accountability in how meaning is interpreted and acted upon. As governments move towards anticipatory and AI-enabled models, the boundary between information and human judgement becomes increasingly important. In a world saturated with data, the defining question is no longer what societies know, but how they determine what matters. This is ultimately a cultural question. Culture provides the shared frameworks through which information acquires significance, priorities are negotiated and collective action becomes possible.

Artificial intelligence can process information at scales impossible for human beings and reveal relationships that would otherwise remain unexplored. A digital twin may reveal that a historic neighbourhood is experiencing demographic change, just as an algorithm may identify a school district facing educational inequities. Yet neither system can determine whether preserving local identity should take precedence over redevelopment, or whether resources should be directed towards equity, excellence or efficiency. Information can reveal conditions; it cannot determine their significance.

As Marcel Mauss observed, societies inherit not only knowledge but also shared frameworks for interpreting reality. Facts do not speak for themselves: their meaning emerges through collective experience, institutional practice and cultural context. This is particularly relevant when decisions concern cultural heritage, education, health or social welfare. A predictive AI model may identify which heritage sites are most vulnerable to deterioration, but it cannot determine why one site should be prioritised over another, especially when communities attribute different forms of historical, symbolic or cultural value to each. Evidence remains essential, but it cannot determine what should be preserved, prioritised or transformed. Such decisions depend on interpretation, responsibility and public legitimacy.

While digital systems are highly effective at preserving information, they are far less effective at replicating the human capacity to interpret information, assign meaning, and exercise judgement.

As governments move towards more anticipatory and AI-enabled models of governance, the distinction between information and meaning becomes increasingly important. Across the world, public institutions are redesigning services around people’s lives rather than administrative structures. Abu Dhabi’s anticipatory-government model, New Zealand's wellbeing framework and the growing adoption of anticipatory governance across OECD countries illustrate some aspects of this evolution.

Yet they also reveal an important boundary: the more routine decisions become automated, the more visible becomes the category of decisions that cannot be automated. Questions concerning life events, community life or cultural continuity cannot be resolved solely through optimisation. They require judgement in determining where automation ends and individual responsibility begins.

The Strategic Value of Human Judgement

Artificial intelligence can expand analytical capabilities, but innovation ultimately depends on human judgement: the capacity to situate information within lived reality, relate facts to values, understand consequences and assume responsibility for action.

The same evolution is taking place in the way institutions approach knowledge. For many years, organisations focused on knowledge management: collecting information, documenting expertise and improving access to data. Increasingly, the challenge concerns knowledge governance: ensuring that information, expertise and experience can be translated into legitimate action.

This distinction matters because not all knowledge can be codified. As Michael Polanyi famously observed, we know more than we can tell. Professional expertise often depends on experience, contextual understanding and practical judgement that cannot be fully reduced to procedures or algorithms. A conservation specialist assessing a historic structure, a teacher responding to the needs of a particular classroom or a physician interpreting a complex diagnosis all draw upon forms of tacit knowledge that emerge through practice rather than formal instruction. Some forms of expertise survive not because they are documented, but because they are transmitted through communities of practice and institutional cultures.

Consequently, the challenge for AI-era institutions is not simply to store more knowledge, but to preserve the conditions through which knowledge supports meaningful judgements.

Culture as Infrastructure for Interpretation

Culture is often discussed through the lens of heritage, participation, tourism or the creative economy. Yet its contribution extends beyond these domains. Just like physical infrastructure supports mobility and digital infrastructure facilitates connectivity, culture enables interpretation. It provides the shared references and values through which societies make sense of change. The capacity to judge is rarely created at the moment a decision is made. Instead, it is accumulated over time through shared experience, collective memory and cultural practice.

For this reason, culture should be understood not only as a sector, but also as an enabling system for governance. It strengthens the interpretative capacity upon which public institutions depend when navigating uncertainty, complexity and competing priorities.

One consequence of this perspective is that cultural measurement must evolve. For decades, cultural indicators focused primarily on participation, attendance, employment and economic contribution. As highlighted by Pope Leo XIV in the Magnifica Humanitas, these metrics almost systematically neglect aspects essential to the overall wellbeing of people and the environment. Thus, the challenge is increasingly to understand how culture contributes to resilience, trust, social cohesion, adaptability and institutional learning. These dimensions are harder to quantify, yet they become valuable in times of technological and social transformation.

Experience from Abu Dhabi illustrates this transition. Building cultural indicators required moving beyond standardised measurement frameworks and engaging with locally relevant forms of value. The objective was not simply to measure cultural activity, but to understand culture's contribution to broader development outcomes and institutional capacity.

Culture as a strategic asset in the age of AI

The transition now underway is often described as a shift towards AI-native institutions. Yet the deeper transformation concerns human judgement itself. As artificial intelligence becomes more pervasive, the less visible cultural capacities that support interpretation, responsibility and legitimacy become more important.

Cultural and educational institutions occupy a particularly important place within this agenda. Their contribution extends beyond the transmission of knowledge or the preservation of heritage. Museums, libraries, archives, schools and universities preserve memory, contextualise information and expose individuals to multiple perspectives and competing narratives. In doing so, they cultivate capacities that become increasingly valuable in an age of artificial intelligence: the ability to exercise judgement, navigate complexity and understand the broader consequences of decisions.

In this context, culture is not simply a beneficiary of innovation: it forms part of the enabling environment through which institutions learn, adapt and maintain legitimacy.

From this perspective, culture becomes a strategic asset because it strengthens the capacity of societies to govern complexity. As artificial intelligence expands the availability of information, the critical challenge lies in transforming information into meaning and meaning into judgement. Yet judgement itself is not the final objective. Public institutions must also secure legitimacy: the capacity to justify decisions, command trust and maintain collective confidence.

Culture provides the interpretative foundations through which this legitimacy is constructed. In this sense, the debate extends beyond AI governance itself. The more capable technological systems become, the more strategic culture becomes as the framework through which societies determine what matters, exercise judgement and translate knowledge into legitimate action.

This insight echoes one of Dostoyevsky’s enduring themes: knowledge and reason alone are insufficient without moral responsibility. As institutions enter an era increasingly shaped by artificial intelligence, the challenge is not only to build smarter systems, but also to strengthen the cultural foundations that enable judgement, legitimacy and responsible action.