AI key terms

The 10 words you need to know to understand artificial intelligence.

AI

(noun)

The ability for machines and computer systems to perform tasks normally requiring human intelligence such as translating between languages, visual perception and decision-making. The ability for AI to solve problems in an uncertain environment is continually improving as it learns from the experiences and data it collects and processes.

Weak AI

(noun)

The ability for machines and computer systems to reproduce intelligent behaviour and handle narrow tasks to solve a specific problem.

Strong AI

(noun)

The ability for machines and computer systems to solve problems by performing tasks in a wide range of fields with an equal or superior level of intelligence to humans. However, its ability to analyse its own reasoning is controversial because it creates the idea of machine “consciousness”.

Augmented intelligence

(noun)

Augmented intelligence comes from the combination of human expertise complemented by the analytical power of artificial intelligence. It enables the analysis of very complex problems, to simulate and predict possible outcomes.

Algorithm

(noun)

A set of instructions given to a machine or computer system to solve a problem in a methodical manner. It enables repetitive processes to be broken down into smaller steps. An adaptive algorithm changes its behaviour or the data it processes by adapting to evolutions in its environment.

Machine learning

(noun)

A learning process that enables a machine to progressively improve how it performs a task without being explicitly programmed. This learning uses statistical techniques or rules built from a variety of data and examples.

Deep learning

(noun)

Part of a broader family of machine learning methods based on learning data representations as opposed to task-specific algorithms. The machine collects and analyses a large amount of information that it can then categorise. Example uses include image recognition and voice technology.

Predictive analytics

(noun)

This encompasses a variety of statistical techniques to produce models that can make predictions about future or hypothetical events. This could be making a range of predictions from customer behaviour through to uncertain or seismic activities. Predictive analytics is made possible by processing large amounts of data (Big Data) combined with the power of AI analytics.

Neural network

(noun)

A computer system composed of interrelated algorithms which have been modelled on biological neurons in the human brain. Thanks to deep learning, these artificial neurons can process information using their experience, and taking into account the situations they’ve already encountered.

Chatbot

(noun)

A dialogue system or conversational interface between a human and a computer program. This “conversational agent” uses AI language analytical capabilities to interpret and respond to user behaviour and conduct a dialogue in the most spontaneous way possible.