AI-generated summary
Artificial Intelligence (AI) lacks a universally agreed-upon definition, with its conceptual roots tracing back to ancient and medieval times. Early ideas, such as Ramón Llull’s 13th-century mechanical reasoning device and ancient Greek automatons, laid foundational thoughts on artificial reasoning. In the mid-20th century, Marvin Minsky described AI as the science of making machines perform tasks requiring human-like intelligence. Today, AI primarily involves algorithms and has seen significant growth due to advances in deep learning and machine learning, which enable machines to analyze vast data sets, recognize patterns, and make decisions with minimal human input. However, the core techniques remain similar to those from decades ago, with progress driven largely by improved computational power.
Despite AI’s advancements, experts highlight its limitations. Pablo Gervás emphasizes that while AI demonstrates isolated intelligent capabilities, it lacks true intelligence because it cannot integrate and switch between skills as humans do. Ramón López de Mántaras points out that no machine has passed the Turing test, indicating AI systems operate based on competence without real comprehension—a view supported by philosopher Daniel Dennett, who argues that AI can read but not understand text. Some argue comprehension is unnecessary since AI’s goal is to mimic intelligent behavior. Neuroscientist Antonio Damasio frames AI as a practical human tool designed to enhance our abilities and evolve our existence, underscoring its role as an extension of human capability rather than an autonomous intelligence.
There is no consensual definition of artificial intelligence. We have asked our experts to answer what is Artificial Intelligence?
There is not an agreed-upon definition of Artificial Intelligence, but many have tried to define it.
Ramón Llull’s (a 13th century, medieval polymath, a herald to the technology) ideas on religion led him to devise a sentient, mechanical wheel that would support or refute arguments, bestowing artificial reasoning on the Christian faith.
Nuria Oliver, during the speech she gave when she accepted her position at the Spanish Royal Academy of Engineering (RAI, by its acronym in Spanish), traced the origins of AI back to the ancient world, with the first automatons that released Hephaestus (the Greek God of fire and metalworking) from his work.
By the mid-20th century, one of the so-called fathers of AI as we know it today, Marvin Minsky, defined it as “the science of making machines do things that would require intelligence if done by humans.”
José Manuel Molina, professor of Computer Science and AI at the Universidad Carlos III de Madrid, points out that AI is basically a set of algorithms (a sequence of steps) executed by a machine.
The current boom in AI is related to deep learning and machine learning. According to SAS, the multinational software company, it automates analytical model building by analyzing large amounts of data to identify patterns and make decisions with minimal human intervention, essentially learning on its own.
However, AI in its current form is limited, and its techniques are the same it had 30 years ago. Only the advances made in infrastructure and computational capacity have made AI able to respond to more needs and have allowed for new applications to emerge. This is why some experts doubt that AI is the right term.
Pablo Gervás, the director of the Natural Interaction based on Language research group at the Instituto Tecnología del Conocimiento (Knowledge Technology Institute) of the Complutense University of Madrid, points out that even though each of AI’s separate capabilities can be considered part of intelligence, they aren’t inherently intelligent. Machines need to develop plenty of capabilities and, above all, know how to combine them and switch from one to the other.
Ramón López de Mántaras asserts that no program or machine has ever passed the Turing test, which evaluates if a machine can exhibit intelligent behavior indistinguishable from that of a human by text. This defines AI as a system based on competence without comprehension, an idea that is widely shared and debated throughout the scientific community. It was first explained by the philosopher of mind and cognitive scientist Daniel Dennett. He believes that, a few years down the road, AI may be able to understand and read a text, but nowadays it can only read it, without any sort of comprehension. Other experts find AI’s ability to comprehend irrelevant, since AI’s purpose is to learn and imitate intelligent behavior.
According to our trustee and famous neuroscientist Antonio Damasio, AI is nothing more than “a practical tool that humans created in order to expand their abilities and improve their existence.” He believes that it is part of how we evolve.
Leverhulme Fellow and Lecturer en University of Cambridge