Artificial intelligence, stupidity and fideism: what digital transition really means

Artificial intelligence, stupidity and fideism: what digital transition really means
Artificial intelligence, stupidity and fideism: what digital transition really means

Environmental crisis, conflicts, inequality. The problem of sustainability of the human species is now impossible to overlook. In this series of posts, as Rethinking Economics Italia we will analyze at the origin the technological, economic and political causes of the problems that afflict modern society and we will provide solutions to combat them.

“Competition means having an advantage both in terms of technology, consequently both in costs and in speed. Digitization helps the group to achieve all this ”.

Words of Peter Oberparleiter, CEO of GKN Powder Metallurgy, introducing them to the Innovation Policy section of the group’s website. When we talk about intelligent manufacturing systems, information circulation networks, dynamism and resilience of integrated production chains, however, it often emerges another side of the coin which represents its transposition into people’s daily lives. In the closing letter of the GKN plant in Florence, with which the dismissal of 422 workers was formalized, there is talk of automotive market contractions such that:

“The organizational structure of the GKN Automotive industrial group reveals itself no longer sustainable, hence the need for immediate action of efficiency, simplification as well as cost reduction”.

It is well understood that many of the jobs canceled by the Covid crisis could only be apparently cyclical. What is structural behind it is increasingly evident. There fourth Industrial Revolution in progress, driven by the digital and technological transition led by Artificial Intelligence (AI) Systems, Information and Communication Technology (ICT) and the new forms of integrated connectivity (5G), to date, still sees the policy and the company struggling in an attempt to bring innovation back towards conditions of environmental and social sustainability.

We live in the enormous contradiction that research is necessary for growth and, at the same time, the present technological frontier risks leaving large sections of the population behind, while a harmonious digital and technological transition must necessarily be sustainable and inclusive. This perspective is threatened on three fronts: social, ethical-economic, environmental.

Generations of workers and technological generations: the need for a new approach to reskilling

The desirable characteristics of a system of Artificial intelligence, as explained by the American computer scientists Peter Norvig e Stuart Russell, that is to think and act rationally, optimizing but also humanly, create obvious problems of social sustainability in how new technologies are driving the restructuring of production processes. In the past, theautomation has been a substitute for many types of jobs, although the bleeding of jobs was absorbed in the long run thanks to the expansion of markets and the creation of new jobs.

Today we are experiencing a phase in which the new generations of the technological frontier follow one another rapidly over the same generation of workers, making a process like innovation of productive processes, socially virtuous, a serious threat to the majority of the population. Specifically, the foresight and understanding of the contexts of Artificial Intelligence technologies, that is what ultimately makes them intelligent, is increasing thanks to the growing availability of Big Data and to the refinement of the processes of Machine Learning: in a rapid way, we are witnessing the transition from Narrow (or Weak) AI systems, technologies that emulate human intelligence on specific tasks (for example, voice assistants), to an increasingly marked development of Artificical General Intelligence (AGI ) which responds in wider contexts to the criteria indicated by Norvig and Russell.

Consequently, the skills requests not to succumb to the technological transition are increasingly refined: the frontier of progress thus increases the segmentation on the labor market, for which workers with high human capital (high-skilled), already paid more on average, are complementary to new technologies and see their position improve, while those who do not belong to this category not only suffer a displacement from the employment point of view, but are also in a position of increasing precariousness .

Innovation, therefore, generates autonomously an unequal distribution of the benefits of change and private investment in the retraining of the workforce (reskilling) is unlikely to be sufficient to fill the gap: a possible line of intervention that harmonizes the transition process in a socially sustainable perspective can be the integration of public policies of reskilling complementary to social security, based on continuous learning. In any case, it must be borne in mind that technologies that increase productivity and crowd out work are inseparable in the short term.

Global power relations, Big Tech bargaining power: the grip on citizens’ rights

L’European Union it is threatened from a geopolitical point of view, since it is experiencing the phase of connection to the development of the transition in a condition of economic subjection with respect to the Asian tigers and the USA. Europe has a share of 13% of the world hardware market and 27% of the market for integrated platforms, while the remainder is almost entirely the preserve of China, Japan, South Korea and the United States.

The same scenario is evident in the global distribution of High Performance Computer, that is, of the current information technology frontier. In Europe we use 33% of the global capacity of computing but we only produce 2%, while in Italy we have about 4% of the processing potential that would be necessary to analyze the amount of data produced daily.

Secondly, we know the current market structure consists of a Big Tech oligopoly destined to expand. Their size has quadrupled in proportion to global GDP from 2000 to 2019 and represents a threat of a not only economic nature: the MIT economist Daron Acemoglu notes how the creation of dominant positions on the market of this entity leads companies to have substantially greater negotiating power than the states, therefore to be able to unilaterally decide the direction of AI development, circumventing both ethical and safety international standards . Acemoglu points out that this is a threat, ultimately, to fundamental human rights, as revealed in the case of the recruitment from Amazon proved discriminatory towards female candidates, and for democracy, as in the recent case of the spread of spyware Pegasus [1][2].

On the need to direct and strengthen research on the basis of social and ethical impact assessments of emerging forms of artificial intelligence, the Artificial Intelligence Act UE of 21/04/21 should be noted: introducing a risk classification, to which it is subject placing AI systems within the Single Market, and establishing new fixed points in the balance of interests between the speed of innovation and public safety, shifting costs and responsibilities to the innovators themselves as much as possible, move in the direction of channeling the development of AI and its implementations in the industrial field along a virtuous path.

Energy sustainability: the convenience of a few drives progress to the detriment of the community

Finally, there is an interconnected problem with the energy issue, which is rarely mentioned. New technologies have an increasing impact on global energy consumption: if today we are at 5%, it is estimated that, in a trend catalyzed by the public and private impulse to digitization, they will be able to absorb about 20% of the electricity available as early as 2030. In the transition phase, investors clearly benefit first mover advantage: ROI (Return On Investment) of large companies for investments in Artificial Intelligence and Big Data is estimated at around 37 (1 dollar makes 37 on average) and this obviously self-feeds crowding of capital which can be observed, with the digital companies which have reached quotation values ​​of around 1000 billion dollars.

The data therefore conceals a final substantial issue of sustainability: the social costs of owning and operating new digital infrastructures are even higher than others production methods and in absolute growth in terms of energy, however private returns do not adequately internalize this aspect and indeed feed it. There is an urgent need for press on energy efficiency of available computing capacity, although this may still not balance the increase in energy consumption driven by increasing usage.

[1] Spyware Pegasus, how it was born, how it spreads and how much it costs, Corriere della Sera 18 July 2021

[2] Pegasus: Spyware sold to governments ‘targets activists’, BBC News, July 19, 2021

PREV Azimut Benetti, turnover of 850 million and orders of 1.2 billion – Corriere.it
NEXT Covid: US health authorities hold back on third vaccine dose – Last Hour