AI in our everyday lives

AI is part of our lives, but what exactly does it mean? We hear about artificial intelligence on television, social media and in newspapers. It is easy to imagine machines that think like us, but that image does not explain how real systems work. Let’s start with a few basic concepts.

Artificial intelligence is present in our lives and will be key in the coming years in economic, political and social progress. Therefore, here we are going to see a simple and understandable definition and characterization of Artificial Intelligence, also called AI, and its different typologies.

Brief conceptualization

Definitions of AI emphasise different things. Broadly, the term covers systems that perform tasks associated with intelligence, such as recognising patterns, solving problems or making decisions. Some use explicit rules; others learn from data drawn from public or private sources. This does not mean that they reproduce the workings of the human brain.

One historical reference is the definition proposed in 2018 by the European Union (EU) “Artificial intelligence (AI) applies to systems that exhibit intelligent behavior, as they are capable of analyzing their environment and taking action – with a certain degree of autonomy – in order to achieve specific objectives” (COM/2018/237 final, Brussels, 25.4.2018). This offers a starting point, although, as we saw in another article, the concept has several interpretations and is easier to understand through examples.

The 1956 Dartmouth workshop was a milestone in the emergence of AI as a discipline. Figures of that period included John McCarthy, Marvin Minsky, Allen Newell, Herbert Simon and Arthur Samuel. Later projects ranged from chess-playing systems to natural-language programs such as Terry Winograd’s SHRDLU, which interacted with a virtual blocks world, and advances in machine learning. At the time this article was published, work by IBM and OpenAI offered familiar examples of that evolution.

Earlier roots can be found in logic and automation. Aristotle studied rules for reaching conclusions through reasoning, while Ctesibius of Alexandria developed automatic regulating mechanisms. These are conceptual and technical precursors, not AI systems in the modern sense.

Different categories of AI

In other articles we have already mentioned some issues about the categorization of AI and its different levels of risk. But here, to better understand what artificial intelligence is, we have to take into account the following categories in which AI can be classified, according to Stuart Russell and Peter Norvig:

1. Systems that think like humans. They try to emulate our thinking and its processes.
2. Systems that act like humans. They aim to reproduce observable human behaviour, such as holding a conversation.
3. Systems that think rationally. They seek conclusions by applying rules of logical reasoning.
4. Systems that act rationally.  They seek actions that achieve their objectives given the information available.

Another useful distinction is between symbolic and subsymbolic approaches. The boundary is not absolute: a system can combine both. Deductive symbolic AI represents knowledge through symbols and explicit rules to draw conclusions. By contrast, subsymbolic AI can learn patterns from data, as neural networks do.

To reflect

AI is a broad field. Machine learning and deep learning are parts of it, not replacements for the term. Another research direction explores automated reasoning: how to connect information and draw conclusions to solve problems. These capabilities do not necessarily mean that a machine reasons like a person. Intriguing, isn’t it?

AI has applications in games, medicine, transport, customer service, finance and industry. It can also appear in everyday settings, from shops to municipal offices. Understanding its capabilities and limits helps us judge when it is useful.

Therefore, if you want to know more about AI and learn what Machine Learning is, take a look at our website to start training in Artificial Intelligence.

References

Gestión (11 November 2018) What is artificial intelligence and what is it for? Available at: https://gestion.pe/tecnologia/inteligencia-artificial-historia-origen-funciona-aplicaciones-categorias-tipos-riesgos-nnda-nnlt-249002-noticia/

López Takeyas, B. (2007) Introduction to Artificial Intelligence. Instituto Tecnológico de Nuevo Laredo. Available at: http://itnuevolaredo.edu.mx/takeyas/Articulos/Inteligencia%20Artificial/ARTICULO%20Introduccion%20a%20la%20Inteligencia%20Artificial.pdf

Singh, A. (09/08/2021) Deep Reasoning: Is this the Next Era of AI? Available at: https://medium.com/@reachtoanamikasingh19/deep-reasoning-is-this-the-next-era-of-ai-4b61b076db27

EU (02/19/2020) Report on the security and liability implications of artificial intelligence, the Internet of Things and robotics Document. Available at: Report on the Security and Civil Liability Impacts of Artificial Intelligence, the Internet of Things and Robotics

Written by

Elena del Castillo

Saturdays.AI

Saturdays.AI is an impact-focused organization on a mission to empower diverse individuals to learn Artificial Intelligence in a collaborative and project-based way, beyond the conventional path of traditional education