The algorithm, the monkey and the mirror: the Machina sapiens shortcut according to Nello Cristianini

Abstract

In 2017 the public debate was revolving around the disorders of the information ecosystem, today the buzzword is ‘artificial intelligence’ (or AI). The conversation is everywhere, but sometimes it remains on the surface, while it would be necessary to reflect on the impact of AI on the disorders of information ecosystem, and, as a consequence, on the librarians who foster AI literacy and who help the communities they serve to understand the context, while they explore new tools.

Two recent books by Nello Cristianini, professor of Artificial Intelligence at the University of Bath, highlight the steps that facilitated the evolution of generative AI and of recommendation systems, though originating some challenges in the process. Generative AI is rooted in the Fifites, in the work of Alan Turing and of the Dartmouth school. It moved from the idea of recreating human intelligence via a logical approach, but later the statistical approach prevailed, where machines learn from data. Big data, from the Internet on, have offered an increasing amount of material about humans, so the machines have got to know them and to mirror them back their image. In the end, the AI is a mirror posing humanity in front of itself, an opportunity for growth.

These technologies offer huge opportunities and some challenges. Greater awareness could help putting the new tools in context, could lead to a critical but constructive approach, to prevent the traps in the infosphere. AI literacy may require different degrees of expertise: top experts must be able to manage machines, but for common citizens  it is enough that they are competent and aware in using available tools. Nello Cristianini’s books help understand how the new machines work, now that they are entering humans’ everyday life.