1.Definition

Artificial intelligence (AI) refers, broadly, to any behaviour of a machine or system that resembles that of a human. In its most basic form, AI relies on programming computers to "reproduce" human actions based on large amounts of data from previously observed similar behaviours.

2. AI Overview

Artificial intelligence is applied to several fields, offering major innovations and their advantages. Among these, we can mention:

•  Healthcare:

Healthcare is a highly promising area for artificial intelligence, currently enhancing the diagnosis, treatment, and prevention of diseases. Key benefits of AI in healthcare include:

-    Improved diagnostics through analysis of medical images.

-    Personalized treatments based on patient data.

-   Disease prevention by identifying at-risk individuals through demographic and lifestyle analysis.

-    Enhanced access to remote care via video consultations.

- Increased productivity by automating administrative and clinical tasks, allowing healthcare professionals to focus on patient care.

•       The finance

Artificial intelligence has the potential to transform the finance sector by automating tasks, improving decision-making, and reducing risk. Key applications of AI in finance include:

- Investment: Analysing financial data to identify opportunities and predict stock and cryptocurrency prices.

- Risk management: Evaluating risk data to protect assets and minimize losses, including fraud and market volatility predictions.

- Compliance: Examining regulatory data to avoid sanctions and enhance company reputations through tools for detecting suspicious transactions.

- Customer service: Providing 24/7 personalized assistance via chatbots, improving customer satisfaction and reducing costs.

-  Administrative automation: Streamlining processes like data entry and account management, freeing employees to focus on higher-value tasks.

•       Logistics and transport

Logistics and transport are key areas for the application of artificial intelligence, improving efficiency and traceability. AI is used to develop autonomous driving systems, optimize traffic management, and improve parcel tracking, inventory management, and demand forecasting. Some major innovations include:

  - Inventory Management: AI analyses data to predict demand, preventing breakages and overstocking.

     - Route Planning: AI evaluates traffic and weather conditions to plan the most efficient routes, reducing costs and improving punctuality.

     - Parcel Tracking: AI facilitates real-time tracking of parcels, increasing transparency and customer satisfaction.

       - Task Automation: AI automates tasks such as data entry and order management.

       - Security: AI detects fraud and security incidents, protecting company assets and data.

     - Autonomous driving systems: These systems make decisions in real time, reducing accidents and facilitating access to transportation for people with disabilities.

    - Connected vehicles: AI allows vehicles to communicate with one another and with infrastructure, enhancing traffic management and providing customized mobility services. This area also includes systems for managing traffic, optimizing transport networks, and improving parcel tracking, inventory management, and demand forecasting, showcasing numerous promising innovations and advantages.

•       Education: 

AI is one of the most innovative areas in education, offering benefits such as personalized learning, automated assessment and improved research.

The main advances of AI in this area include:

          - Personalized learning: AI adapts teaching to individual student needs by analysing their learning data, such as test results and learning preferences.

        - Assessment: Teachers can use AI to set up automated assessments, allowing them to focus on other aspects of teaching while providing more personalized feedback to students.

       -Research: AI tools analyse volumes of data to help researchers better understand education and develop new methods of teaching and learning.

•        Energy and the environment:

Artificial intelligence is used to improve energy efficiency, increase renewable energy production and manage electricity grids.

        - Improving energy efficiency: AI identifies opportunities for improvement in buildings and infrastructure, detecting energy-intensive appliances and developing control algorithms to optimize energy systems.

       - Renewable energy generation and climate forecasting: AI helps predict weather conditions to maximize solar and wind power production. Intelligent sensors also monitor air quality and predict pollution levels.

    - Power management: AI optimizes power distribution and helps respond to outages by analysing consumption data to identify areas of risk.

•        Media:

 AI plays a key role in this area by optimizing content personalization, recommendation, production and delivery.

        -  Content Customization: AI adapts media content to user interests by analysing usage data such as viewing histories and searches.

         -  Content Recommendation: Through the analysis of viewing histories and clicks, AI offers users content that may be of interest to them, which they might not otherwise have found.

      -  Content production: AI automates tasks such as transcription, translation and editing while enabling the creation of innovative formats and interactive experiences.

      - Real-time delivery: Algorithms predict user interests to deliver the appropriate content, customizing the experience based on users’ location and situation.

•       Commerce and marketing:

 AI is key in this area, focusing on personalization, recommendation, automation and analytics.

       - Personalization of the shopping experience** AI adapts the customer experience based on their interests and preferences by analysing purchase data, history and searches.

     - Product Recommendation: Provides products and services tailored to the interests of customers, making it easier to find items they may like.

     - Automation of business tasks: AI simplifies tasks such as appointment making, order management and customer service, allowing employees to focus on higher value-added tasks.

         - Business data analytics: Helps companies better understand their customers and markets by analysing sales, marketing and social media data.

•        Manufacture:

AI focused on efficiency and productivity; This area uses artificial intelligence to automate tasks, improve quality and reduce costs.

          -  Automation of production tasks: AI is used for operations such as assembly and welding, with industrial robots optimizing workflows.

           - Defect Detection: Identifies manufacturing defects before shipment, improving product quality and reducing return costs while helping to design durable products.

           - Safety Monitoring: AI monitors machines and work environments for potential hazards, preventing accidents and improving employee safety.

    - Supply Chain Optimization: helps reduce costs and improve on-time delivery by tracking products throughout their life cycle, which enhances traceability and transparency.

3. How does a Turing machine work? 

Alan Turing invented an abstract machine in 1936 to explain the notion of algorithm, or «mechanical procedure». This machine, intended to perform numerical or symbolic calculations, is the simplest model of this concept. At the time, computers did not yet exist, but his theory laid the foundations of modern computing.

The machine designed by Turing consists of a ribbon divided into boxes where it can write symbols. It can read and write only one box at a time, and moves the ribbon one box to the left or right. The number of possible symbols is limited. For this machine to function as a binary calculator, Turing considers the specific case where the symbols used are 0 and 1. Once the calculation is done, it is on this ribbon that will be written the result of the calculation, the output of the program. At each moment, the ribbon remembers the state of the calculation.

Each program has a description in table form. The machine stops when a state marked as final is reached.

   4. The History of Artificial Intelligence:

The history of artificial intelligence began in 1943 with an article by Warren McCullough and Walter Pitts entitled "A Logical Calculus of Ideas Immanent in Nervous Activity." They proposed the first mathematical model for creating a neural network. In 1950, Marvin Minsky and Dean Edmonds, two Harvard students, created Snarc, the first neural network computer. The same year, Alan Turing published the Turing Test, which remains a benchmark for assessing artificial intelligence. This test lays the foundations of AI by trying to replicate or simulate human intelligence in machines.  In 1956, at the Dartmouth Summer Research Project on Artificial Intelligence, John McCarthy introduced the term ‘’artificial intelligence’’ for the first time. This event, where researchers discuss the vision and goals of AI, is considered by many to be the true starting point for modern artificial intelligence. In 1959, Arthur Samuel invented the term ‘’Machine Learning’’ while working at IBM. In 1989, Yann Le Cun developed the first neural network capable of recognizing handwritten numbers, paving the way for ‘’Deep Learning’’ In 1997, AI marked a historic turning point when IBM’s ‘’Deep Blue’’ system beat world chess champion Gary Kasparov, marking the first victory of a machine over a human.

   5.  AI Milestones:

•           Siri's debut:

A virtual assistant, developed by Apple. Siri understands verbal instructions given by its users and responds to their requests.

•       Alan Turing's test:

The Turing test is a proposed test of artificial intelligence based on a machine's ability to imitate human conversation.

•        AlphaGo defeats Lee Sedol:

Lee Sedol considered the best player in the world in the mid-2000s, and AlphaGo, a go program developed by Google DeepMind; AlphaGo won all except the fourth game.

•  Deep Blue vs. Kasparov:

           IBM's Deep Blue system triumphs over world chess champion Gary Kasparov. For the first time, the machine has defeated Man.

•       GPT;

          ChatGPT is an artificial intelligence capable of generating written content. Developed by OpenAI, is called a generative AI. All you have to do is send it a “prompt” (a question) and GPT executes, with the possibility of adapting its personality.

         6. Artificial intelligence training:

       The training of an artificial intelligence begins with the collection of large data sets. Depending on the type of algorithm used, whether supervised or unsupervised learning, this data is used to identify patterns. During training, errors generated by predictions are evaluated using cost functions. The model then adjusts its internal parameters, via a process, to reduce these errors and improve the accuracy of its predictions.

            7.  The different models of artificial intelligence:

•       Machine learning:

      Machine learning is a subfield of artificial intelligence that allows systems to learn from data and improve their performance without being explicitly programmed for each task. It relies on algorithms that identify patterns in data and use them to make predictions.

•       Deep learning:

      Deep learning is an advanced form of machine learning that uses deep neural networks to process massive amounts of data. It is a key technology for computer vision or Natural Language Processing or NLP applications.

•       Neural networks:

      Artificial neural networks are at the heart of deep learning. Inspired by the human brain, they are composed of several layers of connected neurons that process data in a hierarchical manner. These models are particularly effective at recognizing complex patterns in data, such as images or sounds.

•       LLM (Large Language Models):

      Large Language Models (LLM) are neural networks designed to process large amounts of text data. They are able to understand and generate natural language extremely fluently. Models like GPT and BERT are examples of these, and they are used for tasks like text generation or machine translation.

          8. Developing an ethical artificial intelligence:

   Artificial intelligence raises ethical questions about privacy, especially due to the massive       

 collection of data. It is essential to ensure responsible use of personal information, as in the case of facial recognition technologies, which raise concerns about consent and risks of abuse.

The seven key principles of AI ethics are:

      - Fairness: to avoid discrimination in data sets;

      - Transparency, to make algorithms understandable;

      - No malice: to prevent any negative impact;

 - Responsibility: in development and use; 

- Respect for privacy: with protection of personal data;

- Robustness: to ensure security and resilience of systems; 

- Inclusiveness: to integrate a diversity of views and solve ethical issues