What is artificial intelligence (AI)?
Short answer
Artificial intelligence is the field of building computer systems that perform tasks associated with human intelligence, such as recognising speech or images, translating, predicting and generating text. Most modern AI is machine learning.
AI, machine learning and deep learning
AI is the broad goal. Machine learning (ML) is a way of reaching it: instead of writing rules by hand, engineers train a model on data so it finds patterns itself. Deep learning is a kind of machine learning that uses neural networks with many layers, and it powers most recent advances in vision, speech and language.
Training and inference
Training adjusts a model’s parameters on a large dataset and needs large amounts of computing power, often on specialised chips. Inference is using the finished model to answer a request. Both stages run in data centres, which is why AI is linked to demand for chips, electricity and cloud services.
Narrow and general AI
Today’s systems are "narrow": each is built for particular tasks and can fail in unfamiliar situations. Artificial general intelligence (AGI), a system that matches human ability across most tasks, remains a research goal with no agreed definition or timetable. Generative AI refers to models that produce new text, images, audio or code.
Why it appears in finance news
AI affects company spending plans, semiconductor and cloud demand, regulation and labour markets. Governments and standards bodies, such as the US National Institute of Standards and Technology (NIST), publish frameworks for managing AI risks. Claims about AI capability and market impact vary, so check the source of any specific number.