Skip to content
Permalink
Branch: master
Find file Copy path
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
executable file 80 lines (56 sloc) 6.16 KB

Challenge 2: Technology

Although Artificial Intelligence (AI) is currently not able to reproduce the complex functioning of the human mind [1], but only approximate some limited abilities [2], it is a discipline which, acquired over sixty years of scientific, methodological and technological research, has become pervasive in industry and society. Its models and methods can certainly be conveniently used as tools for the implementation of innovative solutions in complex sociotechnical systems, such as that of public administrations, provided that, together with the opportunities, the limits of their scope of application are understood.

The new frontiers of particular interest for Public Administration are those related to studies and research on how AI systems are able to cooperate in the most effective way with human beings. This approach exploits AI machine learning and adaptation capabilities to provide a humanmachine interaction that best meets the needs of users, their interests and the real context in which they operate. Therefore, it becomes important to conduct research consisting of methodological investigation, modelling, implementation and testing of Artificial Intelligence systems for various domains of interest to PA.

Linguistic technologies (Natural Language Processing - NLP) are certainly of great interest and are the basis of a large number of applications that fall within Artificial Intelligence. The deployment of these technologies, based today to a large extent on Open Source software, requires the availability of specific text datasets (ex. annotated corpora), lexicographical and semantic (ex. wordnet), as well as the dissemination of specialized skills necessary to manage training and adaptation processes to various areas of application (ex. health, justice, finance). The lack or unavailability of adequate resources for the Italian language, together with a lack of skills in the use of NLP technologies, could cause both a loss of competitiveness compared to other nations, and a dependence on platforms and solutions provided by a restricted number of subjects operating under monopoly conditions.

The distinctive characteristics of the “technological challenge” can therefore be identified by two keywords: “personalisation” and “adaptability”.

In fact, the overcoming of this challenge means to be able to create PA systems and services modelled on the multiple needs of citizens, able to evolve with them, able to encourage personalised experiences.

This issue can be approached analytically by linking the most developed sectors and technologies in the field of Artificial Intelligence with the activities and tasks typical of Public Administration.

The areas of use of “intelligent” technologies [3] in Public Administration are innumerable, not only from a long-term perspective, but, in certain cases, even in the current situation.

For example [4] here we can mention:

  • Healthcare system: Diagnostic tools able to assist in the analysis of reports; integration of different sources and data merging; epidemiological analysis to identify public health risks early; instant translation services to facilitate hospital and territorial medical visits to foreigners; predictive tools to evaluate potential risks of disease evolution or to evaluate the effectiveness of therapies; patient assistance tools, able to follow them during treatment; precision medicine, for the identification of personalised treatments; better logistics organisation of healthcare structure activities.
  • Citizen relations: In the simplification of procedures and in order to obtain a two-way communication between PA and citizens and a personalised interaction in which citizens have all the necessary support to satisfy their most varied needs.
  • Judicial system: Simplification of legislation; fraud identification; fight against corruption and crime, especially organised crime; reduction of civil litigation through easier access to legislation and jurisprudence; digitisation of documents and understanding of the text and information present.
  • School system: Automatic evaluation tools; personalisation of teaching material; automated tutoring, by means of recommendation tools to maintain attention; suggestions concerning personalised variations to be introduced in the school programme; extraction of predictive indicators for school drop-out risk
  • Security: AI amplifies the integrated impact of publicly available structured and unstructured data, thanks to which it can support advanced forms of management and prevention in the policing field.
  • Public employment and placement: Organisation of employees and careers; career counselling and management of internal processes and documentation.
  • Mobility and transport: Traffic management, traffic and pollution predictive models, management of public transport logistics, but also autonomous transport solutions; real-time monitoring of sensor data
  • Tax system: Application of AI techniques to identify cases of potential tax avoidance and evasion through the analysis and crossover of data from different sectors of the state.
  • Environmental monitoring: the use of machine learning algorithms on 5G wireless sensor data (ex. video cameras, radioactivity detectors, chemicals, temperature, brightness, humidity, etc.) could allow monitoring and intercepting critical events in the territory (ex. automatic search of events on video surveillance data combined with chemical detection analysis to identify eco-crimes such as spills of harmful substances, similarly possible to define indicators for fires, floods, collapses, etc.).

Footnotes

[1]Research on so-called Full Intelligence and Strong AI, in terms of both Neuro Evolution (NE) and Brain Intelligence (BI), is still in its infancy
[2]Typically, those that do not require more than one second to provide a response to an external stimulus.
[3]Ref. the technologies listed in chapter 2.
[4]Ref. “Challenge reduce inequalities
You can’t perform that action at this time.