Knowledge and Artificial Intelligence: Global ignorance, critical shifts and new potential sources of economic conflict
Gilberto Antonelli · 2026
During the Conference, as well as in the public debate, we often listen comments in line above all with the approach taken by the EU to govern the artificial intelligence (AI) development and the need to make its regulation conditional on the crucial role of human beings, on a values-based approach, on ethical guidelines. The tenet is that AI must remain “human” and linked to “human learning”. All this corresponds to the key words on AI used in the State of the Union address: “guardrail, governance and innovation orientation”. On the other hand, by observing the real evolution of AI, we can notice trends readable through stylized facts, suggestions from the scientific literature, and insights from its most promising research fields, which converge towards a more problematic result. As far as facts are concerned, AI has been defined as “the science and engineering of making intelligent machines”. The basic idea was to teach machines to use ‘human thinking’ starting from a basic model. 1 But this strategy did not lead to good results. This is why it has been substituted with the employment of algorithms generated through machine learning and data mining. This tends to suggest that a sort of creeping competition has been underway between benchmarking ‘knowledge’ or ‘intelligence’. Sometimes, this sort of competition can also take the form of a juxtaposition or antagonism between ‘human thinking’ and ‘statistical thinking’. Looking at the different definitions of the key notions, one can realize that a differentiating component that can be easily sighted in the two definitions is the role assigned to study and experience. As far as scientific trends are concerned, looking at the economic literature, a sharp transition can be observed from an old phase focusing on “knowledge-based economy (KBE)” in the 1980s to a new phase focusing on “data-driven economy (DDE)” launched by OECD in 2011. This transition combined with facts mentioned above, seems to imply that we are not only observing a change in buzzwords and that something deeper is happening. As far as insights coming from promising areas of AI are concerned, a possible support concerning this sort of creeping competition can be found in the ‘literature-based discovery’ (LBD). This involves analysing existing scientific literature, using ChatGPT-style language analysis, to look for new hypotheses, connections or 336 ideas that humans may have missed. LBD is showing good prospects in identifying new experiments to try and even suggesting potential research collaborators: all these functions, we could observe, are normally characteristic of human principal investigators or researchers endowed with their scientific knowledge. On the one side, this can stimulate interdisciplinary work and foster innovation at the boundaries between different scientific fields. On the other, LBD systems can also identify ‘blind spots’ in a given field, possibly linked to omissions, but also opportunistic behaviour of human researchers, and even predict future discoveries and who will make them. If these facts, scientific trends and insights can be deemed robust, at least in the domain of research work, a potential source of conflict could arise between knowledge (incorporated in human researchers) and intelligence (incorporated in machines). Moreover, this conflict could be relevant also for other domains. In this essay we will explore these insights by trying to connect them to two needs. First, the need to resist “digital conformism” by favouring “original thinking”. Second, the need to take into account the lessons on the evolution of the global history of “ignorance” highlighted by Peter Burke in a recent book . SUMMARY: 1. Introduction. – 2. What we know about AI. – 2.1. AI as a technological trajectory. – 2.2. Potential rewards and dangers. – 2.3. Global ignorance and AI. – 2.4. AI as an institutional platform. – 2.4.1. Pillars of AI regulation in EU. – 2.4.2. Production, governance, regulation and innovation. – 3. Insights from the economic literature. – 3.1. Turning from small to vast amounts of micro-data. – 3.2. Turning from knowledge to information and intelligence. – 3.3. Turning from disruption to structural innovation. – 4. Conclusion. – References.