intelligence

General

The capacity to learn, reason, plan, understand and adapt in pursuit of goals, especially in new or unfamiliar situations. No definition is universally accepted: a 2007 compilation by Shane Legg and Marcus Hutter gathered about 70 and distilled them into one line, intelligence as 'an agent's ability to achieve goals in a wide range of environments', and their formal proposal, universal intelligenceuniversal intelligence, n.: A mathematical definition of intelligence as an agent's average ability to achieve goals across every possible environment, with simpler environments counting more., weights performance across all computable environments by their simplicity. In psychology, intelligence is measured by tests whose scores correlate positively across very different tasks, the basis of the g factorg factor, n.: The common factor behind the finding that people who do well on one kind of mental test tend to do well on others. Researchers also look for a similar general factor across AI models.. Disputes concern whether intelligence is one capacity or many, whether it requires understanding or consciousness, and how far it is relative to an environment, a body or a culture.

AI

The property that artificial intelligenceartificial intelligence, n.: Computer systems that do things we would normally credit to a thinking mind, such as holding a conversation, reading an X-ray or planning a route, and the field of research that builds them. research seeks to build and measure in machines. Definitions range from behavioral ones, such as indistinguishability from human performance in the Turing testTuring test, n.: Alan Turing's proposed test of machine intelligence: a judge holds typed conversations with a hidden person and a hidden computer, and the computer passes if the judge cannot consistently pick it out., to measures of skill-acquisition efficiency: François Chollet (2019) defined intelligence as how efficiently a system turns prior knowledge and experience into new skills on unfamiliar tasks, and designed ARC-AGIARC-AGI, prop. n.: Puzzles made of colored grids — and, since 2026, small video-game-like worlds — that people solve easily but AI long found hard, designed to test whether a machine can learn a new skill from just a few examples. to test it. Current systems have uneven profiles, superhuman at some tasks and weak at others that people find easy (jagged intelligencejagged intelligence, n.: The uneven pattern of AI abilities: today's systems can solve very hard problems yet fail at some a child could manage, so their skills form a spiky profile rather than a smooth one.), which complicates any single scale, including proposed levels of AGIlevels of AGI, n.: A scheme from Google DeepMind researchers for grading AI by how well it performs and how broad its skills are, from emerging to superhuman, much as self-driving cars are graded by levels.. Intelligence is conceptually distinct from consciousnessconsciousness, n.: Having experiences: there being something it feels like to be you, as when you see red or feel pain. Whether any AI is or could become conscious is an open and fiercely debated question. and sentiencesentience, n.: The ability to have feelings that are good or bad for you, such as pleasure, pain, comfort or distress. It is the most common test for whether a being's welfare matters morally.: a system could be highly intelligent without experiencing anything.

Benchmark scores measure skills; whether they measure intelligence depends on how much of each skill was bought with training data.