From 1 January this year, theItaly assumed the presidency of the G7 and is committed to placing particular attention to the topic of artificial intelligence (IA). This commitment follows the precedent of Japan, which chaired the G7 in 2023 and initiated the Hiroshima process, culminating in a declaration by political leaders in favor of two documents on the regulation of AI: “Guiding Principles” and the “Code of conduct". During 2023 the topic of artificial intelligence was al center of numerous high-level initiatives, involving a large number of countries. Among these initiatives are the Artificial Intelligence Security Summit organized by the United Kingdom, in which China also participated, and the GPAI ministerial meeting held in New Delhi.
High-level initiatives on Artificial Intelligence (AI) often focus on the risks, but rarely demonstrate its potential. L'Observatory on Italian Public Accounts, led by Giampaolo Galli, instead tried to highlight the many applications of AI, already underway, which have positive impacts on our society. Focusing on the positive aspects of Artificial Intelligence, we seek to promote its adoption as part of the digital transition, convinced that it can significantly contribute to economic productivity, general well-being and the sustainability of welfare systems.
The Observatory then asked itself which one added value can bring the Italian presidency of the G7. According to them, the Italian presidency of the G7 should highlight the positive impacts of AI on collective well-being and economic productivity and also hope for balanced regulation among the G7 countries, avoiding excessive burdens especially for SMEs and startups. It is important that regulation is not only effective but also fair, otherwise it could disincentivize actors who should benefit from the opportunities offered by AI.
What can the G7 do as an Italian presidency?
Il G7 Italian presidency could, explains the Observatory, underline three key concepts regarding Artificial Intelligence (AI).
First, it should promote a regulation that is not too uneven at least among the G7 countries, considering the global nature of the Internet and the challenges businesses face in adhering to different regulations. Harmonization would reduce compliance costs and ensure equal opportunities for consumers and businesses around the world.
Secondly, one complex regulation could be an obstacle for the development of SMEs and represent a barrier to entry for start-ups while it would not be a big problem for large technology companies, which have significant resources to comply with regulations. The “Brussels effect”, highlighted by the GDPR, implies that the rigorous standards introduced in Europe are adopted by large companies and spread globally. This strengthens the market power of non-European big tech, even as Europe takes the lead in regulation.
Finally, it is essential change common perception on Artificial Intelligence as the main source of risks for employment, as the reality is that there will be significant changes in the job market in the coming years. These changes will require active and passive support policies for people, but above all it will be difficult to find workers to fill the available positions.
Artificial intelligence can significantly contribute to productivity growth and general well-being, as well as the sustainability of welfare systems. This is particularly relevant for countries with high public debts and low birth rates, such as Italy and Japan.
Let's see, according to the Italian Public Accounts Observatory, what the useful applications and positive aspects that technology already offers in various sectors.
Artificial Intelligence in everyday life
Artificial intelligence already exists an integral part of our daily life, with several applications and devices that we use regularly:
- Satellite navigators: Like Google Maps, they use AI algorithms to calculate optimal routes and monitor traffic in real time.
- AI cars and planes: Modern cars use computer vision systems for lane keeping and pedestrian detection, while flight automation assists pilots in different phases of flight.
- Suggestors, proofreaders and translators: They are present on PCs and mobile phones, providing word suggestions, corrections and automatic translations.
- Personalized searches: Search engines offer personalized results based on user preferences.
- Targeted Advertising: Uses AI algorithms to tailor ads to users' interests.
- Photo Management: Automatically recognizes faces, objects and environments in photos and provides personalized suggestions.
- Voice assistants: Like Siri and Alexa, they understand natural language and can perform various tasks.
- AI Chatbot: How Chat GPT, offer assistance and automatic responses
Artificial intelligence in industry and agriculture
A McKinsey report lists the business functions which can be made more efficient with the use of AI including demand forecasting, raw material purchasing and inventory management. Two concrete examples inindustry they come from two large companies such as Barilla and Leonardo that use AI to optimize production and management processes.
- Barilla: Use Computer Vision for quality control and optimization of production processes.
- Digital Twin by Leonardo: A project that uses AI models to simulate and predict the behavior of real systems.
In 'agriculture, Artificial Intelligence has various applications, from experiments to common practices. It is used for improve yield e predict phenomena such as parasite attacks, commodity prices or suggest the sowing date.
Artificial intelligence in the public dimension
Artificial Intelligence in Public Administration offers numerous opportunities for improve efficiency and reduce costs. Here are some examples:
- E-government in Estonia: Estonia has implemented a number of advanced digital services that facilitate the use of AI. With e-Tax, the Estonian tax system is almost entirely digitalised, with 98% of tax returns filed digitally, enabling the use of AI to combat tax evasion. X-Road is a platform that enables the secure exchange of data between government agencies and the private sector, facilitating the adoption of AI to streamline practices. The customer service service, active via AI chatbot since 2018, resolved approximately half of the requests, saving time and resources for operators.
- Altext Citibeats in Ireland: Altext Citibeats uses AI to collect information on citizens' proposals. In Ireland, thousands of tweets from Dublin residents are analyzed monthly to understand community opinions and ensure efficient use of public resources.
- Other uses of AI in PA: In addition to the cases mentioned above, AI is used for the management of flows and levels of city water, the speeding up of processes for obtaining documents such as those relating to immigration, the identification of critical points in response plans to natural disasters and the prediction of financial distress in Italian municipalities, offering valuable tools for monitoring authorities such as the Court of Auditors.
In Health, AI is used in different areas, from diagnosis to biomedical research:
- FDA-Enabled Devices in the US: There are currently 171 medical devices using AI and Machine Learning (AI/ML) techniques approved by the FDA. These devices cover a wide range of applications, from radiology to diabetes management and cardiopulmonary care. For example, there are smartphone glucometers that offer continuous and precise monitoring of blood sugar levels and electronic stethoscopes that improve diagnostic accuracy in cardiology.
- Analysis of anatomical structures or lesions: In radiology and cardiology, AI is used to locate and analyze anatomical structures or lesions in diagnostic images. Platforms like Circle Cardiovascular Imaging's "cvi42" use AI algorithms to improve the diagnosis and quantification of cardiovascular disease, enabling more efficient use of healthcare resources.
- Cardiac conditions: AI is also used for the analysis of electrocardiographic (ECG) data, using deep-learning models to identify cardiac pathologies and phenotypic characteristics, improving the diagnosis and monitoring of patients.
- Biomedical research: AlphaFold is a significant example of how AI is revolutionizing biomedical research. This system can predict the three-dimensional structure of proteins with unprecedented precision, enabling a deeper understanding of diseases and potential therapies.
In 'education, AI is instead playing an increasingly important role in the field of education, offering a series of tools and applications aimed at improving teaching and learning.
- EduChat: This model, developed by East China Normal University, provides services for teaching and learning. Its codes, data and parameters are shared as open source, allowing for wide use and customization.
- MathGPT: Developed by the TAL Education Group, MathGPT is a Language Learning Model (LLM) focused on solving mathematical problems, offering support to students in tackling complex mathematical challenges.
- Intelligent Tutoring Systems: There are several AI applications that provide personalized feedback to students, automatic grading, inattention detection, and educational resource suggestions. For example, IBM Watson Discovery offers learning path personalization and learning analytics.
- Predicting academic failure: AI can be used to predict student academic failure, allowing educators to proactively intervene to provide personalized support strategies and improve student outcomes.
- Photomath: This application assists students in studying mathematics by allowing them to take a photo of the problem and receive detailed steps to solve it. Photomath helps students understand the process of solving mathematical problems.
Artificial intelligence in mobility and infrastructure
Artificial intelligence is also revolutionizing the mobility and urban infrastructure sector, offering innovative solutions for more efficient and sustainable management of cities. Some examples are:
- Smart cities: AI is used in smart cities to acquire and interpret environmental information in order to act rationally. Applications include mobility, logistics and infrastructure management.
- Smart grid: Smart grids, powered by AI, ensure consumption efficiency and cost minimization by managing consumer demand. Initiatives like Enel in Italy use AI algorithms to predict grid outages and optimize energy production.
- Urban lighting: “Smart” solutions for urban lighting allow remote control and monitoring of various environmental factors. Projects like “Humble Lamppost” introduce smart streetlights to Europe.
- Air traffic: Tools such as Enav's Arrival Manager improve air traffic management by reducing delays and fuel consumption thanks to better management of approaching aircraft.
- Car sharing: Car rental services, especially electric and car sharing, take advantage of AI to inform users about consumption and emissions, contributing to the ecosystem. Companies like Clem in France are implementing these solutions.
- Smart parking: The intelligent management of parking areas, through diversified access rights, reduces congestion and pollution. Companies such as Passport in the USA and Cleverciti in Germany are developing innovative solutions in this sector.
