Strolling through informatics #28 – Digital transformation needs a different perspective

previous #27 ––– next #29

by Enrico Nardelli

(versione italiana qua)

In recent posts we have been discussing various aspects of the relationship between informatics and the world of work, first reflecting on the need for cultural maturation based on teaching informatics from compulsory school onwards, then on the role played by informatics in increasing productivity and finally on how, for ordinary people, it is necessary to develop transparent and user-friendly computer systems. Let's now tackle the issue of a proper vision of what we call "digital transformation" which is nothing other than computer automation.

Typically, in fact, an organization acquires a computer system to replace, through an automated system, cognitive faculties previously exercised by one or more people. The functions to be automated are identified and replaced by a computer system, that is, a "cognitive machine", which is the corresponding element in digital society to the traditional machine of industrial society.

This substitution, like any automation, occurs to improve productivity, that is, to increase output or decrease costs or both. So far so good: work automation has been for centuries the key factor ensuring constant productivity growth. Having acquired a computer system therefore means having replaced one or more people with one or more "cognitive machines". But these, without adaptive capacity, are unable to evolve to cope with changing contextual conditions. This is why the acquisition or development of any computer application must follow a different path.

It is necessary to change first and foremost the mental paradigm with which computer automation is approached. Every organization knows well, when hiring an economist, an engineer, a lawyer or an accountant, that what that person can do at the beginning will not remain unchanged over time, but will evolve, because the person will learn on the job a whole series of details relevant to the organization itself and will adapt their behavior as their operational scenario evolves. Initially under the guidance of their supervisor, and then with an increasingly greater degree of autonomy. If part of this cognitive work is transferred to computer systems, this flexibility and capacity for evolution, which are specific to and characteristic of human beings, are lost. Not understanding this perspective means continuing to waste money on developing computer systems.

Before computerizing a work process, it must be analyzed in depth and we must understand how its automation impacts work organization, internal power relationships and external ones. When we think we can outsource IT services, saving on personnel, we always end up discovering that we have spent more to adapt these services brought outside the organization to every minor change in the surrounding reality, modifications that an IT professional employed internally would have known how to handle in a fraction of the time.

Unfortunately, we still cannot build systems for transforming information and knowledge with a level of reliability, predictability and security comparable to what traditional engineering artifacts offer. We are still at the level of building medieval cathedrals, with the difference that if a cathedral collapsed there were fewer consequences for society than the collapse of a computer system, given our almost total dependence on their proper functioning. We still don't know how to build systems that have a semantic understanding of reality and its changes at the level of an elementary school child, despite what appears to us from interaction with generative artificial intelligence systems (GAI). More and more frequently, the digital tools we use daily turn into mechanisms to control and record what we say and do, then transform this data into opportunities for enormous profit for the few, leaving us crumbs of personal utility. In short, informatics that could and should be used to make us better off becomes an instrument of oppression and stress.

The approach to use, therefore, is to consider a digital transformation process as the acquisition of a certain number of people with certain basic skills. No personnel recruiter expects to always find "the perfect candidate", because this is not the norm at all. The aim is to find a person with a profile good enough to be able to "take the field" effectively and then, from there, evolve.

With computer systems to be developed for digital transformation, the same approach must be adopted. This doesn't mean taking the first system that comes along, but making the incremental and co-constructed development (by users and developers, by clients and suppliers) of the system itself part of the acquisition process. Exactly as happens with employees. All those who deal with these issues know how complicated it is to integrate a team of 10 employees into a group of 100, all the more so the greater the cognitive and non-physical component of the activities carried out in the organization. When you digitize a business process you are essentially doing the same thing. Why should we proceed differently? If we do so, it's because we haven't understood that informatics is a radically different automation from any other and requires a different approach.

We will explore this further in the next post.

[[The posts in this series are based on the Author's book (in Italian) La rivoluzione informatica: conoscenza, consapevolezza e potere nella società digitale, (= The Informatics Revolution: Knowledge, Awareness and Power in the Digital Society) to which readers are referred for further reading]].

--
The original version (in italian) has been published by "Osservatorio sullo Stato digitale" (= Observatory on Digital State) of IRPA - Istituto di Ricerche sulla Pubblica Amministrazione (= Research Institute on Public Administration) on 16 April 2025.