Artificial intelligence: definition, history and applications

ChatGPT summary of this article:

‘This article gives an introduction to artificial intelligence (AI) and explains the different approaches to AI, including machine learning, deep learning and natural language processing. It explains how AI is being used in various fields such as healthcare, finance and industry. It also highlights the benefits of using AI, such as increasing the efficiency and accuracy of business processes, as well as the challenges associated with AI, such as model complexity and privacy and security concerns. Finally, the article encourages businesses to explore the opportunities offered by AI and develop a clear strategy for integrating AI into their operations.’

‘The creation of artificial intelligence would be the greatest event in human history. But it could also be the ultimate.’
Stephen Hawking.

What is artificial intelligence?

Defined as the process of imitating human intelligence, artificial intelligence (or AI) is based on the development of algorithms designed to give computers the ability to think and interact like human beings.

History of artificial intelligence

Machines designed to simulate human behaviour were imagined long before the beginnings of robotics and computing. However, these were automata, whose actions and reactions in a well-defined context had to be thought out and integrated into complex mechanisms beforehand.

The history of artificial intelligence began in 1943 with the publication of the article ‘A Logical Calculus of Ideas Immanent in Nervous Activity’ by Warren McCullough and Walter Pitts. This article describes the first mathematical model to create a neural network. In 1950, the first neural network computer was developed by Marvin Minsky and Dean Edmonds, followed by the famous Turing test to evaluate artificial intelligence models.

The term ‘artificial intelligence’ was first used in 1956 by John McCarthy at the Dartmouth Summer Research Project on Artificial Intelligence. This conference is considered by many researchers to be the birth of artificial intelligence and data science.

However, artificial intelligence was to face phases of doubt on the part of the scientific community, and many AI research projects were cancelled. This is what is known as the first and second AI winters, which lasted until the 1990s.

The technological advances of the 2000s, however, gave a glimpse of a revival in artificial intelligence, and investment in AI research projects resumed. In 2016, Google's AlphaGo model beat Lee Sedol, one of the world's best Go players in his own discipline.

Today, artificial intelligence is used by many companies in all sectors, for a multitude of applications. AI is constantly progressing and surprising us with its performance.

Artificial Intelligence

A. Definition of Artificial Intelligence

Today, human beings and machines produce data faster than it is possible to ingest and interpret it in order to make decisions. Artificial intelligence is the answer to this problem and represents the future of complex decision-making processes.

But this process of human intelligence being harnessed by computers is based on several successive phases:

• Machine learning, i.e. understanding the information and the rules for using it.
• Reasoning, during which the algorithm uses the rules to draw conclusions.
• Self-correction

B. Applications of Artificial Intelligence

Artificial Intelligence applications can be divided into distinct categories according to the type of data they manipulate or the way they learn:

1. Machine Learning: Machine learning is a sub-domain of artificial intelligence, enabling algorithms to learn and improve their performance based on the data they receive.
2. Deep learning: A sub-branch of machine learning, deep learning is based on a stack of layers of neural networks. Its aim is to be able to imitate the actions of the human brain using neural networks.
3. Computer Vision: Enabling machines to recognise, identify and classify images, computer vision is the AI equivalent of the human eye and our brain's ability to process and analyse perceived images.
4. NLP (Natural language processing): Mainly used in translation systems such as Google Translate or Deepl, as well as voice assistants such as Siri or Alexa. NLP or automatic language processing enables machines to understand, translate or generate spoken or written human language.
   
   
   

C. Cas d’usages

Find out how artificial intelligence technologies can support your business through these projects carried out by Theodo Data & AI:

1. Predictive maintenance: As part of an assessment of its poles, the Colas Rail design office teamed up with Sicara to automate one of the stages of the project: the 3D drawing of the poles generated by LIDAR and then its 2D integration into internal business software. This stage, which used to take 15 minutes to complete manually, has now been reduced to 45 seconds.
2. Recommendation engine: As part of a project for the Culture Pass, Sicara developed an algorithm capable of recommending cultural events to young people under the age of 18, based on their centres of interest.
3. Predictive API: For a major retail brand, Sicara has developed a predictive API POC that can infer a person's wrinkle level from their photo.
4. Automatic invoicing: For an industrial player in the corporate catering sector, we built an application for automatic invoicing of meal trays using algorithms that are at the cutting edge of visual recognition research.
   
   If you too have an artificial intelligence project or questions about the various use cases, don't hesitate to contact us. Our teams will be delighted to help you develop complex solutions!