Artificial intelligence (AI), in its broadest sense, is

intelligence Intelligence has been defined in many ways: the capacity for abstraction, logic, understanding, self-awareness, learning, emotional knowledge, reasoning, planning, creativity, critical thinking, and problem-solving. It can be described as the a ...

exhibited by machines, particularly

computer systems A computer is a machine that can be programmed to carry out sequences of arithmetic or logical operations (computation) automatically. Modern digital electronic computers can perform generic sets of operations known as programs. These program ...

. It is a

field of research An academic discipline or academic field is a subdivision of knowledge that is taught and researched at the college or university level. Disciplines are defined (in part) and recognized by the academic journals in which research is published, an ...

computer science Computer science is the study of computation, automation, and information. Computer science spans theoretical disciplines (such as algorithms, theory of computation, information theory, and automation) to practical disciplines (includin ...

that develops and studies methods and

software Software is a set of computer programs and associated software documentation, documentation and data (computing), data. This is in contrast to Computer hardware, hardware, from which the system is built and which actually performs the work. ...

that enable machines to perceive their environment and use learning and intelligence to take actions that maximize their chances of achieving defined goals. Such machines may be called AIs. High-profile

applications of AI Artificial intelligence (AI) has been used in applications throughout industry and academia. In a manner analogous to electricity or computers, AI serves as a general-purpose technology. AI programes emulate perception and understanding, and are ...

include advanced

web search engine A search engine is a software system designed to carry out web searches. They search the World Wide Web in a systematic way for particular information specified in a textual web search query. The search results are generally presented in a ...

s (e.g.,

Google Search Google Search (also known simply as Google) is a search engine provided by Google. Handling more than 3.5 billion searches per day, it has a 92% share of the global search engine market. It is also the most-visited website in the world. The ...

);

recommendation systems A recommender system, or a recommendation system (sometimes replacing 'system' with a synonym such as platform or engine), is a subclass of information filtering system that provide suggestions for items that are most pertinent to a particular u ...

(used by

YouTube YouTube is a global online video sharing and social media platform headquartered in San Bruno, California. It was launched on February 14, 2005, by Steve Chen, Chad Hurley, and Jawed Karim. It is owned by Google, and is the second most ...

Amazon Amazon most often refers to: * Amazons, a tribe of female warriors in Greek mythology * Amazon rainforest, a rainforest covering most of the Amazon basin * Amazon River, in South America * Amazon (company), an American multinational technolog ...

, and

Netflix Netflix, Inc. is an American subscription video on-demand over-the-top streaming service and production company based in Los Gatos, California. Founded in 1997 by Reed Hastings and Marc Randolph in Scotts Valley, California, it offers a ...

);

virtual assistants Virtual may refer to: * Virtual (horse), a thoroughbred racehorse * Virtual channel, a channel designation which differs from that of the actual radio channel (or range of frequencies) on which the signal travels * Virtual function, a programming ...

(e.g.,

Google Assistant Google Assistant is a virtual assistant software application developed by Google that is primarily available on mobile and home automation devices. Based on artificial intelligence, Google Assistant can engage in two-way conversations, unlike t ...

Siri Siri ( ) is a virtual assistant that is part of Apple Inc.'s iOS, iPadOS, watchOS, macOS, tvOS, and audioOS operating systems. It uses voice queries, gesture based control, focus-tracking and a natural-language user interface to answer ques ...

, and

Alexa Alexa may refer to: Technology *Amazon Alexa, a virtual assistant developed by Amazon * Alexa Internet, a defunct website ranking and traffic analysis service * Arri Alexa, a digital motion picture camera People *Alexa (name), a given name and ...

);

autonomous vehicles Vehicular automation involves the use of mechatronics, artificial intelligence, and multi-agent systems to assist the operator of a vehicle (car, aircraft, watercraft, or otherwise).Hu, J.; Bhowmick, P.; Lanzon, A.,Group Coordinated Control ...

(e.g.,

Waymo Waymo LLC, formerly known as the Google self-driving car project, is an American autonomous driving technology company headquartered in Mountain View, California. It is a subsidiary of Alphabet Inc, the parent company of Google. Waymo oper ...

);

generative Generative may refer to: * Generative actor, a person who instigates social change * Generative art, art that has been created using an autonomous system that is frequently, but not necessarily, implemented using a computer * Generative music, ...

and

creative Creative may refer to: *Creativity, phenomenon whereby something new and valuable is created * "Creative" (song), a 2008 song by Leon Jackson * Creative class, a proposed socioeconomic class * Creative destruction, an economic term * Creative dir ...

tools (e.g.,

ChatGPT ChatGPT (Generative Pre-trained Transformer) is a chatbot launched by OpenAI in November 2022. It is built on top of OpenAI's GPT-3 family of large language models, and is fine-tuned (an approach to transfer learning) with both supervised and ...

and

AI art Artificial intelligence art is any artwork created through the use of artificial intelligence. Tools and processes Imagery There are many mechanisms for creating AI art, including procedural 'rule-based' generation of images using mathemat ...

); and

superhuman The term superhuman refers to humans or human-like beings with enhanced qualities and abilities that exceed those naturally found in humans. These qualities may be acquired through natural ability, self-actualization or technological aids. Th ...

play and analysis in

strategy game A strategy game or strategic game is a game (e.g. a board game) in which the players' uncoerced, and often autonomous, decision-making skills have a high significance in determining the outcome. Almost all strategy games require internal decisio ...

s (e.g.,

chess Chess is a board game for two players, called White and Black, each controlling an army of chess pieces in their color, with the objective to checkmate the opponent's king. It is sometimes called international chess or Western chess to dist ...

and Go). However, many AI applications are not perceived as AI: "A lot of cutting edge AI has filtered into general applications, often without being called AI because once something becomes useful enough and common enough it's not labeled AI anymore." Various subfields of AI research are centered around particular goals and the use of particular tools. The traditional goals of AI research include

reasoning Reason is the capacity of consciously applying logic by drawing conclusions from new or existing information, with the aim of seeking the truth. It is closely associated with such characteristically human activities as philosophy, science, lang ...

knowledge representation Knowledge representation and reasoning (KRR, KR&R, KR²) is the field of artificial intelligence (AI) dedicated to representing information about the world in a form that a computer system can use to solve complex tasks such as diagnosing a medic ...

planning Planning is the process of thinking regarding the activities required to achieve a desired goal. Planning is based on foresight, the fundamental capacity for mental time travel. The evolution of forethought, the capacity to think ahead, is c ...

, learning,

natural language processing Natural language processing (NLP) is an interdisciplinary subfield of linguistics, computer science, and artificial intelligence concerned with the interactions between computers and human language, in particular how to program computers to proc ...

, perception, and support for

robotics Robotics is an interdisciplinarity, interdisciplinary branch of computer science and engineering. Robotics involves design, construction, operation, and use of robots. The goal of robotics is to design machines that can help and assist human ...

General intelligence The ''g'' factor (also known as general intelligence, general mental ability or general intelligence factor) is a construct developed in psychometric investigations of cognitive abilities and human intelligence. It is a variable that summarizes ...

—the ability to complete any task performed by a human on an at least equal level—is among the field's long-term goals. To reach these goals, AI researchers have adapted and integrated a wide range of techniques, including

search Searching or search may refer to: Computing technology * Search algorithm, including keyword search ** :Search algorithms * Search and optimization for problem solving in artificial intelligence * Search engine technology, software for findi ...

and

mathematical optimization Mathematical optimization (alternatively spelled ''optimisation'') or mathematical programming is the selection of a best element, with regard to some criterion, from some set of available alternatives. It is generally divided into two subfi ...

formal logic Logic is the study of correct reasoning. It includes both formal and informal logic. Formal logic is the science of deductively valid inferences or of logical truths. It is a formal science investigating how conclusions follow from premise ...

artificial neural network Artificial neural networks (ANNs), usually simply called neural networks (NNs) or neural nets, are computing systems inspired by the biological neural networks that constitute animal brains. An ANN is based on a collection of connected units ...

s, and methods based on statistics,

operations research Operations research ( en-GB, operational research) (U.S. Air Force Specialty Code: Operations Analysis), often shortened to the initialism OR, is a discipline that deals with the development and application of analytical methods to improve dec ...

, and

economics Economics () is the social science that studies the production, distribution, and consumption of goods and services. Economics focuses on the behaviour and interactions of economic agents and how economies work. Microeconomics analy ...

. AI also draws upon

psychology Psychology is the scientific study of mind and behavior. Psychology includes the study of conscious and unconscious phenomena, including feelings and thoughts. It is an academic discipline of immense scope, crossing the boundaries betwe ...

linguistics Linguistics is the scientific study of human language. It is called a scientific study because it entails a comprehensive, systematic, objective, and precise analysis of all aspects of language, particularly its nature and structure. Lingu ...

, philosophy,

neuroscience Neuroscience is the science, scientific study of the nervous system (the brain, spinal cord, and peripheral nervous system), its functions and disorders. It is a Multidisciplinary approach, multidisciplinary science that combines physiology, an ...

, and other fields. Artificial intelligence was founded as an academic discipline in 1956, and the field went through multiple cycles of optimism, followed by periods of disappointment and loss of funding, known as

AI winter In the history of artificial intelligence, an AI winter is a period of reduced funding and interest in artificial intelligence research. Funding and interest vastly increased after 2012 when deep learning outperformed previous AI techniques. This growth accelerated further after 2017 with the

transformer architecture A transformer is a deep learning architecture developed by researchers at Google and based on the multi-head attention mechanism, proposed in a 2017 paper " Attention Is All You Need". Text is converted to numerical representations called toke ...

, and by the early 2020s hundreds of billions of dollars were being invested in AI (known as the "

AI boom The AI boom, or AI spring, is the ongoing period of rapid progress in the field of artificial intelligence. Prominent examples include protein folding prediction and generative AI, led by laboratories including Google DeepMind and OpenAI. ...

"). The widespread use of AI in the 21st century exposed several unintended consequences and harms in the present and raised concerns about its risks and long-term effects in the future, prompting discussions about regulatory policies to ensure the safety and benefits of the technology.

Goals

The general problem of simulating (or creating) intelligence has been broken into subproblems. These consist of particular traits or capabilities that researchers expect an intelligent system to display. The traits described below have received the most attention and cover the scope of AI research.

Reasoning and problem-solving

Early researchers developed algorithms that imitated step-by-step reasoning that humans use when they solve puzzles or make logical deductions. By the late 1980s and 1990s, methods were developed for dealing with uncertain or incomplete information, employing concepts from

probability Probability is the branch of mathematics concerning numerical descriptions of how likely an Event (probability theory), event is to occur, or how likely it is that a proposition is true. The probability of an event is a number between 0 and ...

and

. Many of these algorithms are insufficient for solving large reasoning problems because they experience a "combinatorial explosion": They become exponentially slower as the problems grow. Intractability and efficiency and the

combinatorial explosion In mathematics, a combinatorial explosion is the rapid growth of the complexity of a problem due to how the combinatorics of the problem is affected by the input, constraints, and bounds of the problem. Combinatorial explosion is sometimes used to ...

: Even humans rarely use the step-by-step deduction that early AI research could model. They solve most of their problems using fast, intuitive judgments.Psychological evidence of the prevalence of sub-symbolic reasoning and knowledge: , , , Accurate and efficient reasoning is an unsolved problem.

Knowledge representation

Knowledge representation Knowledge representation and reasoning (KRR, KR&R, KR²) is the field of artificial intelligence (AI) dedicated to representing information about the world in a form that a computer system can use to solve complex tasks such as diagnosing a medic ...

and

knowledge engineering Knowledge engineering (KE) refers to all technical, scientific and social aspects involved in building, maintaining and using knowledge-based systems. Background Expert systems One of the first examples of an expert system was MYCIN, an appl ...

allow AI programs to answer questions intelligently and make deductions about real-world facts. Formal knowledge representations are used in content-based indexing and retrieval, scene interpretation, clinical decision support, knowledge discovery (mining "interesting" and actionable inferences from large

database In computing, a database is an organized collection of data stored and accessed electronically. Small databases can be stored on a file system, while large databases are hosted on computer clusters or cloud storage. The design of databases spa ...

s), and other areas. A

knowledge base A knowledge base (KB) is a technology used to store complex structured and unstructured information used by a computer system. The initial use of the term was in connection with expert systems, which were the first knowledge-based systems. ...

is a body of knowledge represented in a form that can be used by a program. An

ontology In metaphysics, ontology is the philosophical study of being, as well as related concepts such as existence, becoming, and reality. Ontology addresses questions like how entities are grouped into categories and which of these entities ...

is the set of objects, relations, concepts, and properties used by a particular domain of knowledge. Knowledge bases need to represent things such as objects, properties, categories, and relations between objects; situations, events, states, and time; causes and effects; knowledge about knowledge (what we know about what other people know);

default reasoning Default logic is a non-monotonic logic proposed by Raymond Reiter to formalize reasoning with default assumptions. Default logic can express facts like “by default, something is true”; by contrast, standard logic can only express that somethin ...

(things that humans assume are true until they are told differently and will remain true even when other facts are changing);

Default reasoning Default logic is a non-monotonic logic proposed by Raymond Reiter to formalize reasoning with default assumptions. Default logic can express facts like “by default, something is true”; by contrast, standard logic can only express that somethin ...

Frame problem In artificial intelligence, the frame problem describes an issue with using first-order logic (FOL) to express facts about a robot in the world. Representing the state of a robot with traditional FOL requires the use of many axioms that simply impl ...

, default logic,

non-monotonic logic A non-monotonic logic is a formal logic whose conclusion relation is not monotonic. In other words, non-monotonic logics are devised to capture and represent defeasible inferences (cf. defeasible reasoning), i.e., a kind of inference in which re ...

circumscription Circumscription may refer to: *Circumscribed circle * Circumscription (logic) *Circumscription (taxonomy) *Circumscription theory The circumscription theory is a theory of the role of warfare in state formation in political anthropology, created ...

closed world assumption The closed-world assumption (CWA), in a formal system of logic used for knowledge representation, is the presumption that a statement that is true is also known to be true. Therefore, conversely, what is not currently known to be true, is false. Th ...

abduction Abduction may refer to: Media Film and television * "Abduction" (''The Outer Limits''), a 2001 television episode * " Abduction" (''Death Note'') a Japanese animation television series * " Abductions" (''Totally Spies!''), a 2002 episode of an ...

: , , , (Poole ''et al.'' places abduction under "default reasoning". Luger ''et al.'' places this under "uncertain reasoning"). and many other aspects and domains of knowledge. Among the most difficult problems in knowledge representation are the breadth of commonsense knowledge (the set of atomic facts that the average person knows is enormous);Breadth of commonsense knowledge: , , , (

qualification problem In philosophy Philosophy (from , ) is the systematized study of general and fundamental questions, such as those about existence, reason, Epistemology, knowledge, Ethics, values, Philosophy of mind, mind, and Philosophy of language, langu ...

) and the sub-symbolic form of most commonsense knowledge (much of what people know is not represented as "facts" or "statements" that they could express verbally). There is also the difficulty of knowledge acquisition, the problem of obtaining knowledge for AI applications.

Planning and decision-making

An "agent" is anything that perceives and takes actions in the world. A

rational agent A rational agent or rational being is a person or entity that always aims to perform optimal actions based on given premises and information. A rational agent can be anything that makes decisions, typically a person, firm, machine, or software. ...

has goals or preferences and takes actions to make them happen. In

automated planning Automation describes a wide range of technologies that reduce human intervention in processes, namely by predetermining decision criteria, subprocess relationships, and related actions, as well as embodying those predeterminations in machines ...

, the agent has a specific goal. In

automated decision-making Automated decision-making (ADM) involves the use of data, machines and algorithms to make decisions in a range of contexts, including public administration, business, health, education, law, employment, transport, media and entertainment, with var ...

, the agent has preferences—there are some situations it would prefer to be in, and some situations it is trying to avoid. The decision-making agent assigns a number to each situation (called the "

utility As a topic of economics, utility is used to model worth or value. Its usage has evolved significantly over time. The term was introduced initially as a measure of pleasure or happiness as part of the theory of utilitarianism by moral philosoph ...

") that measures how much the agent prefers it. For each possible action, it can calculate the "

expected utility The expected utility hypothesis is a popular concept in economics that serves as a reference guide for decisions when the payoff is uncertain. The theory recommends which option rational individuals should choose in a complex situation, based on the ...

": the

of all possible outcomes of the action, weighted by the probability that the outcome will occur. It can then choose the action with the maximum expected utility. In classical planning, the agent knows exactly what the effect of any action will be. In most real-world problems, however, the agent may not be certain about the situation they are in (it is "unknown" or "unobservable") and it may not know for certain what will happen after each possible action (it is not "deterministic"). It must choose an action by making a probabilistic guess and then reassess the situation to see if the action worked. In some problems, the agent's preferences may be uncertain, especially if there are other agents or humans involved. These can be learned (e.g., with

inverse reinforcement learning Reinforcement learning (RL) is an area of machine learning concerned with how intelligent agents ought to take actions in an environment in order to maximize the notion of cumulative reward. Reinforcement learning is one of three basic machine ...

), or the agent can seek information to improve its preferences.

Information value theory Value of information (VOI or VoI) is the amount a decision maker would be willing to pay for information prior to making a decision. Similar terms VoI is sometimes distinguished into value of perfect information, also called value of clairvoyance ( ...

can be used to weigh the value of exploratory or experimental actions. The space of possible future actions and situations is typically

intractably In theoretical computer science and mathematics, computational complexity theory focuses on classifying computational problems according to their resource usage, and relating these classes to each other. A computational problem is a task solved b ...

large, so the agents must take actions and evaluate situations while being uncertain of what the outcome will be. A Markov decision process has a transition model that describes the probability that a particular action will change the state in a particular way and a

reward function Reinforcement learning (RL) is an area of machine learning concerned with how intelligent agents ought to take actions in an environment in order to maximize the notion of cumulative reward. Reinforcement learning is one of three basic machine ...

that supplies the utility of each state and the cost of each action. A

policy Policy is a deliberate system of guidelines to guide decisions and achieve rational outcomes. A policy is a statement of intent and is implemented as a procedure or protocol. Policies are generally adopted by a governance body within an orga ...

associates a decision with each possible state. The policy could be calculated (e.g., by

iteration Iteration is the repetition of a process in order to generate a (possibly unbounded) sequence of outcomes. Each repetition of the process is a single iteration, and the outcome of each iteration is then the starting point of the next iteration. ...

), be

heuristic A heuristic (; ), or heuristic technique, is any approach to problem solving or self-discovery that employs a practical method that is not guaranteed to be optimal, perfect, or rational, but is nevertheless sufficient for reaching an immediat ...

, or it can be learned. Game theory describes the rational behavior of multiple interacting agents and is used in AI programs that make decisions that involve other agents.

Learning

Machine learning Machine learning (ML) is a field of inquiry devoted to understanding and building methods that 'learn', that is, methods that leverage data to improve performance on some set of tasks. It is seen as a part of artificial intelligence. Machine ...

is the study of programs that can improve their performance on a given task automatically. It has been a part of AI from the beginning. Supervised and unsupervised learning

There are several kinds of machine learning.

Unsupervised learning Unsupervised learning is a type of algorithm that learns patterns from untagged data. The hope is that through mimicry, which is an important mode of learning in people, the machine is forced to build a concise representation of its world and t ...

analyzes a stream of data and finds patterns and makes predictions without any other guidance.

Supervised learning Supervised learning (SL) is a machine learning paradigm for problems where the available data consists of labelled examples, meaning that each data point contains features (covariates) and an associated label. The goal of supervised learning alg ...

requires labeling the training data with the expected answers, and comes in two main varieties:

classification Classification is a process related to categorization, the process in which ideas and objects are recognized, differentiated and understood. Classification is the grouping of related facts into classes. It may also refer to: Business, organizat ...

(where the program must learn to predict what category the input belongs in) and

regression Regression or regressions may refer to: Science * Marine regression, coastal advance due to falling sea level, the opposite of marine transgression * Regression (medicine), a characteristic of diseases to express lighter symptoms or less extent ( ...

(where the program must deduce a numeric function based on numeric input).

: (Definition), (Techniques) In

reinforcement learning Reinforcement learning (RL) is an area of machine learning concerned with how intelligent agents ought to take actions in an environment in order to maximize the notion of cumulative reward. Reinforcement learning is one of three basic machine ...

, the agent is rewarded for good responses and punished for bad ones. The agent learns to choose responses that are classified as "good".

Transfer learning Transfer learning (TL) is a research problem in machine learning (ML) that focuses on storing knowledge gained while solving one problem and applying it to a different but related problem. For example, knowledge gained while learning to recognize ...

is when the knowledge gained from one problem is applied to a new problem. Deep learning is a type of machine learning that runs inputs through biologically inspired artificial neural networks for all of these types of learning.

Computational learning theory In computer science, computational learning theory (or just learning theory) is a subfield of artificial intelligence devoted to studying the design and analysis of machine learning algorithms. Overview Theoretical results in machine learning m ...

can assess learners by

computational complexity In computer science, the computational complexity or simply complexity of an algorithm is the amount of resources required to run it. Particular focus is given to computation time (generally measured by the number of needed elementary operations) ...

, by

sample complexity The sample complexity of a machine learning algorithm represents the number of training-samples that it needs in order to successfully learn a target function. More precisely, the sample complexity is the number of training-samples that we need to ...

(how much data is required), or by other notions of

optimization Mathematical optimization (alternatively spelled ''optimisation'') or mathematical programming is the selection of a best element, with regard to some criterion, from some set of available alternatives. It is generally divided into two subfi ...

Natural language processing

Natural language processing Natural language processing (NLP) is an interdisciplinary subfield of linguistics, computer science, and artificial intelligence concerned with the interactions between computers and human language, in particular how to program computers to proc ...

(NLP) allows programs to read, write and communicate in human languages such as

English English usually refers to: * English language * English people English may also refer to: Peoples, culture, and language * ''English'', an adjective for something of, from, or related to England ** English national id ...

. Specific problems include

speech recognition Speech recognition is an interdisciplinary subfield of computer science and computational linguistics that develops methodologies and technologies that enable the recognition and translation of spoken language into text by computers with the ma ...

speech synthesis Speech synthesis is the artificial production of human speech. A computer system used for this purpose is called a speech synthesizer, and can be implemented in software or hardware products. A text-to-speech (TTS) system converts normal languag ...

machine translation Machine translation, sometimes referred to by the abbreviation MT (not to be confused with computer-aided translation, machine-aided human translation or interactive translation), is a sub-field of computational linguistics that investigates t ...

information extraction Information extraction (IE) is the task of automatically extracting structured information from unstructured and/or semi-structured machine-readable documents and other electronically represented sources. In most of the cases this activity concer ...

, information retrieval and

question answering Question answering (QA) is a computer science discipline within the fields of information retrieval and natural language processing (NLP), which is concerned with building systems that automatically answer questions posed by humans in a natural l ...

. Early work, based on

Noam Chomsky Avram Noam Chomsky (born December 7, 1928) is an American public intellectual: a linguist, philosopher, cognitive scientist, historian, social critic, and political activist. Sometimes called "the father of modern linguistics", Chomsky is ...

generative grammar Generative grammar, or generativism , is a linguistic theory that regards linguistics as the study of a hypothesised innate grammatical structure. It is a biological or biologistic modification of earlier structuralist theories of linguistic ...

and

semantic network A semantic network, or frame network is a knowledge base that represents semantic relations between concepts in a network. This is often used as a form of knowledge representation. It is a directed or undirected graph consisting of vertices ...

s, had difficulty with

word-sense disambiguation Word-sense disambiguation (WSD) is the process of identifying which sense of a word is meant in a sentence or other segment of context. In human language processing and cognition, it is usually subconscious/automatic but can often come to cons ...

unless restricted to small domains called " micro-worlds" (due to the common sense knowledge problem).

Margaret Masterman Margaret Masterman (4 May 1910 – 1 April 1986) was a British linguist and philosopher, most known for her pioneering work in the field of computational linguistics and especially machine translation. She founded the Cambridge Language R ...

believed that it was meaning and not grammar that was the key to understanding languages, and that

thesauri A thesaurus (plural ''thesauri'' or ''thesauruses'') or synonym dictionary is a reference work for finding synonyms and sometimes antonyms of words. They are often used by writers to help find the best word to express an idea: Synonym dictionar ...

and not dictionaries should be the basis of computational language structure. Modern deep learning techniques for NLP include

word embedding In natural language processing (NLP), word embedding is a term used for the representation of words for text analysis, typically in the form of a real-valued vector that encodes the meaning of the word such that the words that are closer in the v ...

(representing words, typically as vectors encoding their meaning),

transformer A transformer is a passive component that transfers electrical energy from one electrical circuit to another circuit, or multiple circuits. A varying current in any coil of the transformer produces a varying magnetic flux in the transformer' ...

s (a deep learning architecture using an

attention Attention is the behavioral and cognitive process of selectively concentrating on a discrete aspect of information, whether considered subjective or objective, while ignoring other perceivable information. William James (1890) wrote that "Att ...

mechanism), and others. In 2019, generative pre-trained transformer (or "GPT") language models began to generate coherent text, and by 2023, these models were able to get human-level scores on the

bar exam A bar examination is an examination administered by the bar association of a jurisdiction that a lawyer must pass in order to be admitted to the bar of that jurisdiction. Australia Administering bar exams is the responsibility of the bar associa ...

SAT The SAT ( ) is a standardized test widely used for college admissions in the United States. Since its debut in 1926, its name and scoring have changed several times; originally called the Scholastic Aptitude Test, it was later called the Schol ...

test,

GRE The Graduate Record Examinations (GRE) is a standardized test that is an admissions requirement for many graduate schools in the United States and Canada and a few other countries. The GRE is owned and administered by Educational Testing Servi ...

test, and many other real-world applications.

Perception

Machine perception Machine perception is the capability of a computer system to interpret data in a manner that is similar to the way humans use their senses to relate to the world around them. The basic method that the computers take in and respond to their environ ...

is the ability to use input from sensors (such as cameras, microphones, wireless signals, active

lidar Lidar (, also LIDAR, or LiDAR; sometimes LADAR) is a method for determining ranges (variable distance) by targeting an object or a surface with a laser and measuring the time for the reflected light to return to the receiver. It can also be ...

, sonar, radar, and

tactile sensor A tactile sensor is a device that measures information arising from physical interaction with its environment. Tactile sensors are generally modeled after the biological sense of cutaneous touch which is capable of detecting stimuli resultin ...

s) to deduce aspects of the world.

Computer vision Computer vision is an Interdisciplinarity, interdisciplinary scientific field that deals with how computers can gain high-level understanding from digital images or videos. From the perspective of engineering, it seeks to understand and automate t ...

is the ability to analyze visual input. The field includes

image classification Computer vision is an interdisciplinary scientific field that deals with how computers can gain high-level understanding from digital images or videos. From the perspective of engineering, it seeks to understand and automate tasks that the human ...

facial recognition Facial recognition or face recognition may refer to: * Face detection, often a step done before facial recognition * Face perception, the process by which the human brain understands and interprets the face * Pareidolia, which involves, in part, se ...

object recognition Object recognition – technology in the field of computer vision for finding and identifying objects in an image or video sequence. Humans recognize a multitude of objects in images with little effort, despite the fact that the image of the ...

, object tracking, and

robotic perception Robotic sensing is a subarea of robotics science intended to give robots sensing capabilities. Robotic sensing mainly gives robots the ability to see,Roh SG, Choi HR (Jan 2009).3-D Tag-Based RFID System for Recognition of Object" IEEE Transaction ...

Social intelligence

Affective computing Affective computing is the study and development of systems and devices that can recognize, interpret, process, and simulate human affects. It is an interdisciplinary field spanning computer science, psychology, and cognitive science. While so ...

is an interdisciplinary umbrella that comprises systems that recognize, interpret, process, or simulate human feeling, emotion, and mood. For example, some

virtual assistant An intelligent virtual assistant (IVA) or intelligent personal assistant (IPA) is a software agent that can perform tasks or services for an individual based on commands or questions. The term " chatbot" is sometimes used to refer to virtua ...

s are programmed to speak conversationally or even to banter humorously; it makes them appear more sensitive to the emotional dynamics of human interaction, or to otherwise facilitate

human–computer interaction Human–computer interaction (HCI) is research in the design and the use of computer technology, which focuses on the interfaces between people ( users) and computers. HCI researchers observe the ways humans interact with computers and design ...

. However, this tends to give naïve users an unrealistic conception of the intelligence of existing computer agents. Moderate successes related to affective computing include textual

sentiment analysis Sentiment analysis (also known as opinion mining or emotion AI) is the use of natural language processing, text analysis, computational linguistics, and biometrics to systematically identify, extract, quantify, and study affective states and subjec ...

and, more recently,

multimodal sentiment analysis Multimodal sentiment analysis is a new dimension of the traditional text-based sentiment analysis, which goes beyond the analysis of texts, and includes other modalities such as audio and visual data. It can be bimodal, which includes different com ...

, wherein AI classifies the effects displayed by a videotaped subject.

General intelligence

A machine with

artificial general intelligence Artificial general intelligence (AGI) is the ability of an intelligent agent to understand or learn any intellectual task that a human being can. It is a primary goal of some artificial intelligence research and a common topic in science fict ...

should be able to solve a wide variety of problems with breadth and versatility similar to

human intelligence Human intelligence is the intellectual capability of humans, which is marked by complex cognitive feats and high levels of motivation and self-awareness. High intelligence is associated with better outcomes in life. Through intelligence, humans ...

Artificial general intelligence Artificial general intelligence (AGI) is the ability of an intelligent agent to understand or learn any intellectual task that a human being can. It is a primary goal of some artificial intelligence research and a common topic in science fict ...

:
Proposal for the modern version:
Warnings of overspecialization in AI from leading researchers: , ,

Techniques

AI research uses a wide variety of techniques to accomplish the goals above.

Search and optimization

AI can solve many problems by intelligently searching through many possible solutions. There are two very different kinds of search used in AI:

state space search State space search is a process used in the field of computer science, including artificial intelligence (AI), in which successive configurations or ''states'' of an instance are considered, with the intention of finding a ''goal state'' with the ...

and local search.

State space search

State space search State space search is a process used in the field of computer science, including artificial intelligence (AI), in which successive configurations or ''states'' of an instance are considered, with the intention of finding a ''goal state'' with the ...

searches through a tree of possible states to try to find a goal state. For example,

algorithms search through trees of goals and subgoals, attempting to find a path to a target goal, a process called means-ends analysis. Simple exhaustive searches are rarely sufficient for most real-world problems: the search space (the number of places to search) quickly grows to astronomical numbers. The result is a search that is

too slow High five is a friendly gesture in which one individual slaps another's hand. High five (and variants such as Hi5, Hi-5, and Hi-Five) may also refer to: Music * Hi-5 (Australian group), an Australian children's musical group * Hi-5 (Greek band), ...

or never completes. "

Heuristics A heuristic (; ), or heuristic technique, is any approach to problem solving or self-discovery that employs a practical method that is not guaranteed to be optimal, perfect, or rational, but is nevertheless sufficient for reaching an immediate, ...

" or "rules of thumb" can help prioritize choices that are more likely to reach a goal.

Adversarial search In computer science, a search algorithm is an algorithm designed to solve a search problem. Search algorithms work to retrieve information stored within particular data structure, or calculated in the search space of a problem domain, with eith ...

is used for game-playing programs, such as chess or Go. It searches through a

tree In botany, a tree is a perennial plant with an elongated stem, or trunk, usually supporting branches and leaves. In some usages, the definition of a tree may be narrower, including only woody plants with secondary growth, plants that are ...

of possible moves and countermoves, looking for a winning position.

Local search

Local search uses

to find a solution to a problem. It begins with some form of guess and refines it incrementally.

Gradient descent In mathematics, gradient descent (also often called steepest descent) is a first-order iterative optimization algorithm for finding a local minimum of a differentiable function. The idea is to take repeated steps in the opposite direction of ...

is a type of local search that optimizes a set of numerical parameters by incrementally adjusting them to minimize a

loss function In mathematical optimization and decision theory, a loss function or cost function (sometimes also called an error function) is a function that maps an event or values of one or more variables onto a real number intuitively representing some "co ...

. Variants of

gradient descent In mathematics, gradient descent (also often called steepest descent) is a first-order iterative optimization algorithm for finding a local minimum of a differentiable function. The idea is to take repeated steps in the opposite direction of ...

are commonly used to train neural networks. Another type of local search is

evolutionary computation In computer science, evolutionary computation is a family of algorithms for global optimization inspired by biological evolution, and the subfield of artificial intelligence and soft computing studying these algorithms. In technical terms, ...

, which aims to iteratively improve a set of candidate solutions by "mutating" and "recombining" them, selecting only the fittest to survive each generation. Distributed search processes can coordinate via

swarm intelligence Swarm intelligence (SI) is the collective behavior of decentralized, self-organized systems, natural or artificial. The concept is employed in work on artificial intelligence. The expression was introduced by Gerardo Beni and Jing Wang in 1989, ...

algorithms. Two popular swarm algorithms used in search are particle swarm optimization (inspired by bird

flocking A flock is a large group of animals, especially birds, sheep, or goats. Flock or flocking also may refer to: Computing * Flock (messaging service), a communication app for teams * Flock (web browser), a discontinued web browser * Flock system ca ...

) and

ant colony optimization In computer science and operations research, the ant colony optimization algorithm (ACO) is a probabilistic technique for solving computational problems which can be reduced to finding good paths through graphs. Artificial ants stand for multi ...

(inspired by

ant trail Ants are eusocial insects of the family Formicidae and, along with the related wasps and bees, belong to the order Hymenoptera. Ants evolved from vespoid wasp ancestors in the Cretaceous period. More than 13,800 of an estimated total of 22 ...

s).

Logic

Formal

logic Logic is the study of correct reasoning. It includes both formal and informal logic. Formal logic is the science of deductively valid inferences or of logical truths. It is a formal science investigating how conclusions follow from premis ...

is used for

and

. Formal logic comes in two main forms:

propositional logic Propositional calculus is a branch of logic. It is also called propositional logic, statement logic, sentential calculus, sentential logic, or sometimes zeroth-order logic. It deals with propositions (which can be true or false) and relations ...

(which operates on statements that are true or false and uses

logical connective In logic, a logical connective (also called a logical operator, sentential connective, or sentential operator) is a logical constant. They can be used to connect logical formulas. For instance in the syntax of propositional logic, the binary ...

s such as "and", "or", "not" and "implies") and

predicate logic First-order logic—also known as predicate logic, quantificational logic, and first-order predicate calculus—is a collection of formal systems used in mathematics, philosophy, linguistics, and computer science. First-order logic uses quanti ...

(which also operates on objects, predicates and relations and uses quantifiers such as "''Every'' ''X'' is a ''Y''" and "There are ''some'' ''X''s that are ''Y''s").

Deductive reasoning Deductive reasoning is the mental process of drawing deductive inferences. An inference is deductively valid if its conclusion follows logically from its premises, i.e. if it is impossible for the premises to be true and the conclusion to be false ...

in logic is the process of proving a new statement ( conclusion) from other statements that are given and assumed to be true (the

premise A premise or premiss is a true or false statement that helps form the body of an argument, which logically leads to a true or false conclusion. A premise makes a declarative statement about its subject matter which enables a reader to either agre ...

s). Proofs can be structured as proof

trees In botany, a tree is a perennial plant with an elongated stem, or trunk, usually supporting branches and leaves. In some usages, the definition of a tree may be narrower, including only woody plants with secondary growth, plants that are ...

, in which nodes are labelled by sentences, and children nodes are connected to parent nodes by

inference rule In the philosophy of logic, a rule of inference, inference rule or transformation rule is a logical form consisting of a function which takes premises, analyzes their syntax, and returns a conclusion (or conclusions). For example, the rule of ...

s. Given a problem and a set of premises, problem-solving reduces to searching for a proof tree whose root node is labelled by a solution of the problem and whose

leaf nodes In computer science, a tree is a widely used abstract data type that represents a hierarchical tree structure with a set of connected nodes. Each node in the tree can be connected to many children (depending on the type of tree), but must be c ...

are labelled by premises or

axiom An axiom, postulate, or assumption is a statement that is taken to be true, to serve as a premise or starting point for further reasoning and arguments. The word comes from the Ancient Greek word (), meaning 'that which is thought worthy o ...

s. In the case of

Horn clause In mathematical logic and logic programming, a Horn clause is a logical formula of a particular rule-like form which gives it useful properties for use in logic programming, formal specification, and model theory. Horn clauses are named for the logi ...

s, problem-solving search can be performed by reasoning forwards from the premises or backwards from the problem. In the more general case of the clausal form of

first-order logic First-order logic—also known as predicate logic, quantificational logic, and first-order predicate calculus—is a collection of formal systems used in mathematics, philosophy, linguistics, and computer science. First-order logic uses quanti ...

resolution Resolution(s) may refer to: Common meanings * Resolution (debate), the statement which is debated in policy debate * Resolution (law), a written motion adopted by a deliberative body * New Year's resolution, a commitment that an individual ma ...

is a single, axiom-free rule of inference, in which a problem is solved by proving a contradiction from premises that include the negation of the problem to be solved. Inference in both Horn clause logic and first-order logic is undecidable, and therefore

intractable Intractable may refer to: * Intractable conflict, a form of complex, severe, and enduring conflict * Intractable pain, pain which cannot be controlled/cured by any known treatment * Intractable problem In theoretical computer science and mathema ...

. However, backward reasoning with Horn clauses, which underpins computation in the

logic programming Logic programming is a programming paradigm which is largely based on formal logic Logic is the study of correct reasoning. It includes both formal and informal logic. Formal logic is the science of deductively valid inferences or of log ...

language

Prolog Prolog is a logic programming language associated with artificial intelligence and computational linguistics. Prolog has its roots in first-order logic, a formal logic, and unlike many other programming languages, Prolog is intended primarily a ...

, is

Turing complete Alan Mathison Turing (; 23 June 1912 – 7 June 1954) was an English mathematician, computer scientist, logician, cryptanalyst, philosopher, and theoretical biologist. Turing was highly influential in the development of theoretical co ...

. Moreover, its efficiency is competitive with computation in other

symbolic programming In computer programming, symbolic programming is a programming paradigm in which the program can manipulate its own formulas and program components as if they were plain data. Through symbolic programming, complex processes can be developed that b ...

languages.

Fuzzy logic Fuzzy logic is a form of many-valued logic in which the truth value of variables may be any real number between 0 and 1. It is employed to handle the concept of partial truth, where the truth value may range between completely true and complet ...

assigns a "degree of truth" between 0 and 1. It can therefore handle propositions that are vague and partially true.

Non-monotonic logic A non-monotonic logic is a formal logic whose conclusion relation is not monotonic. In other words, non-monotonic logics are devised to capture and represent defeasible inferences (cf. defeasible reasoning), i.e., a kind of inference in which re ...

s, including logic programming with

negation as failure Negation as failure (NAF, for short) is a non-monotonic inference rule in logic programming, used to derive \mathrm~p (i.e. that ~p is assumed not to hold) from failure to derive ~p. Note that \mathrm ~p can be different from the statement \neg p ...

, are designed to handle

. Other specialized versions of logic have been developed to describe many complex domains.

Probabilistic methods for uncertain reasoning

Many problems in AI (including in reasoning, planning, learning, perception, and robotics) require the agent to operate with incomplete or uncertain information. AI researchers have devised a number of tools to solve these problems using methods from

theory and economics.Stochastic methods for uncertain reasoning: , , , Precise mathematical tools have been developed that analyze how an agent can make choices and plan, using

decision theory Decision theory (or the theory of choice; not to be confused with choice theory) is a branch of applied probability theory concerned with the theory of making decisions based on assigning probabilities to various factors and assigning numerical ...

decision analysis Decision analysis (DA) is the discipline comprising the philosophy, methodology, and professional practice necessary to address important decisions in a formal manner. Decision analysis includes many procedures, methods, and tools for identifying, ...

, and

information value theory Value of information (VOI or VoI) is the amount a decision maker would be willing to pay for information prior to making a decision. Similar terms VoI is sometimes distinguished into value of perfect information, also called value of clairvoyance ( ...

. These tools include models such as Markov decision processes, dynamic

decision network Decision may refer to: Law and politics * Judgment (law), as the outcome of a legal case *Landmark decision, the outcome of a case that sets a legal precedent * ''Per curiam'' decision, by a court with multiple judges Books * ''Decision'' (nove ...

s, game theory and

mechanism design Mechanism design is a field in economics and game theory that takes an objectives-first approach to designing economic mechanisms or incentives, toward desired objectives, in strategic settings, where players act rationally. Because it starts a ...

Bayesian network A Bayesian network (also known as a Bayes network, Bayes net, belief network, or decision network) is a probabilistic graphical model that represents a set of variables and their conditional dependencies via a directed acyclic graph (DAG). Bay ...

s are a tool that can be used for

(using the Bayesian inference algorithm), learning (using the

expectation–maximization algorithm In statistics, an expectation–maximization (EM) algorithm is an iterative method to find (local) maximum likelihood or maximum a posteriori (MAP) estimates of parameters in statistical models, where the model depends on unobserved latent variab ...

(using

s) and

perception Perception () is the organization, identification, and interpretation of sensory information in order to represent and understand the presented information or environment. All perception involves signals that go through the nervous system, ...

(using dynamic Bayesian networks). Probabilistic algorithms can also be used for filtering, prediction, smoothing, and finding explanations for streams of data, thus helping

systems analyze processes that occur over time (e.g.,

hidden Markov model A hidden Markov model (HMM) is a statistical Markov model in which the system being modeled is assumed to be a Markov process — call it X — with unobservable ("''hidden''") states. As part of the definition, HMM requires that there be an ob ...

s or

Kalman filter For statistics and control theory, Kalman filtering, also known as linear quadratic estimation (LQE), is an algorithm that uses a series of measurements observed over time, including statistical noise and other inaccuracies, and produces estima ...

s).Stochastic temporal models:

Hidden Markov model A hidden Markov model (HMM) is a statistical Markov model in which the system being modeled is assumed to be a Markov process — call it X — with unobservable ("''hidden''") states. As part of the definition, HMM requires that there be an ob ...

s: Dynamic Bayesian networks: EM_Clustering_of_Old_Faithful_data

Classifiers and statistical learning methods

The simplest AI applications can be divided into two types: classifiers (e.g., "if shiny then diamond"), on one hand, and controllers (e.g., "if diamond then pick up"), on the other hand. Classifiers are functions that use

pattern matching In computer science, pattern matching is the act of checking a given sequence of tokens for the presence of the constituents of some pattern. In contrast to pattern recognition, the match usually has to be exact: "either it will or will not be ...

to determine the closest match. They can be fine-tuned based on chosen examples using

supervised learning Supervised learning (SL) is a machine learning paradigm for problems where the available data consists of labelled examples, meaning that each data point contains features (covariates) and an associated label. The goal of supervised learning alg ...

. Each pattern (also called an "

observation Observation is the active acquisition of information from a primary source. In living beings, observation employs the senses. In science, observation can also involve the perception and recording of data via the use of scientific instruments. Th ...

") is labeled with a certain predefined class. All the observations combined with their class labels are known as a

data set A data set (or dataset) is a collection of data. In the case of tabular data, a data set corresponds to one or more database tables, where every column of a table represents a particular variable, and each row corresponds to a given record of the d ...

. When a new observation is received, that observation is classified based on previous experience. There are many kinds of classifiers in use. The

decision tree A decision tree is a decision support tool that uses a tree-like model of decisions and their possible consequences, including chance event outcomes, resource costs, and utility. It is one way to display an algorithm that only contains co ...

is the simplest and most widely used symbolic machine learning algorithm.

K-nearest neighbor In statistics, the ''k''-nearest neighbors algorithm (''k''-NN) is a non-parametric supervised learning method first developed by Evelyn Fix and Joseph Hodges in 1951, and later expanded by Thomas Cover. It is used for classification and regre ...

algorithm was the most widely used analogical AI until the mid-1990s, and

Kernel methods In machine learning, kernel machines are a class of algorithms for pattern analysis, whose best known member is the support-vector machine (SVM). The general task of pattern analysis is to find and study general types of relations (for example c ...

such as the

support vector machine In machine learning, support vector machines (SVMs, also support vector networks) are supervised learning models with associated learning algorithms that analyze data for classification and regression analysis. Developed at AT&T Bell Laboratories ...

(SVM) displaced k-nearest neighbor in the 1990s. The

naive Bayes classifier In statistics, naive Bayes classifiers are a family of simple " probabilistic classifiers" based on applying Bayes' theorem with strong (naive) independence assumptions between the features (see Bayes classifier). They are among the simplest Baye ...

is reportedly the "most widely used learner" at Google, due in part to its scalability.

Neural networks A neural network is a network or circuit of biological neurons, or, in a modern sense, an artificial neural network, composed of artificial neurons or nodes. Thus, a neural network is either a biological neural network, made up of biological ...

are also used as classifiers.

Artificial neural networks

An artificial neural network is based on a collection of nodes also known as

artificial neurons An artificial neuron is a mathematical function conceived as a model of biological neurons, a neural network. Artificial neurons are elementary units in an artificial neural network. The artificial neuron receives one or more inputs (representing ...

, which loosely model the

neurons A neuron, neurone, or nerve cell is an electrically excitable cell that communicates with other cells via specialized connections called synapses. The neuron is the main component of nervous tissue in all animals except sponges and placozoa. ...

in a biological brain. It is trained to recognise patterns; once trained, it can recognise those patterns in fresh data. There is an input, at least one hidden layer of nodes and an output. Each node applies a function and once the

weight In science and engineering, the weight of an object is the force acting on the object due to gravity. Some standard textbooks define weight as a vector quantity, the gravitational force acting on the object. Others define weight as a scalar q ...

crosses its specified threshold, the data is transmitted to the next layer. A network is typically called a deep neural network if it has at least 2 hidden layers.Neural networks: , Learning algorithms for neural networks use local search to choose the weights that will get the right output for each input during training. The most common training technique is the

backpropagation In machine learning, backpropagation (backprop, BP) is a widely used algorithm for training feedforward artificial neural networks. Generalizations of backpropagation exist for other artificial neural networks (ANNs), and for functions gener ...

algorithm. Neural networks learn to model complex relationships between inputs and outputs and find patterns in data. In theory, a neural network can learn any function. In

feedforward neural network A feedforward neural network (FNN) is an artificial neural network wherein connections between the nodes do ''not'' form a cycle. As such, it is different from its descendant: recurrent neural networks. The feedforward neural network was the ...

s the signal passes in only one direction.

Recurrent neural network A recurrent neural network (RNN) is a class of artificial neural networks where connections between nodes can create a cycle, allowing output from some nodes to affect subsequent input to the same nodes. This allows it to exhibit temporal dynamic ...

s feed the output signal back into the input, which allows short-term memories of previous input events.

Long short term memory Long short-term memory (LSTM) is an artificial neural network used in the fields of artificial intelligence and deep learning. Unlike standard feedforward neural networks, LSTM has feedback connections. Such a recurrent neural network (RNN) ...

is the most successful network architecture for recurrent networks.

Perceptron In machine learning, the perceptron (or McCulloch-Pitts neuron) is an algorithm for supervised classification, supervised learning of binary classification, binary classifiers. A binary classifier is a function which can decide whether or not an ...

s use only a single layer of neurons; deep learning uses multiple layers.

Convolutional neural network In deep learning, a convolutional neural network (CNN, or ConvNet) is a class of artificial neural network (ANN), most commonly applied to analyze visual imagery. CNNs are also known as Shift Invariant or Space Invariant Artificial Neural Netwo ...

s strengthen the connection between neurons that are "close" to each other—this is especially important in

image processing An image is a visual representation of something. It can be two-dimensional, three-dimensional, or somehow otherwise feed into the visual system to convey information. An image can be an artifact, such as a photograph or other two-dimension ...

, where a local set of neurons must identify an "edge" before the network can identify an object.

Deep learning

Deep learning Deep learning: , , , uses several layers of neurons between the network's inputs and outputs. The multiple layers can progressively extract higher-level features from the raw input. For example, in

, lower layers may identify edges, while higher layers may identify the concepts relevant to a human such as digits, letters, or faces. Deep learning has profoundly improved the performance of programs in many important subfields of artificial intelligence, including

computer vision Computer vision is an Interdisciplinarity, interdisciplinary scientific field that deals with how computers can gain high-level understanding from digital images or videos. From the perspective of engineering, it seeks to understand and automate t ...

, and others. The reason that deep learning performs so well in so many applications is not known as of 2023. The sudden success of deep learning in 2012–2015 did not occur because of some new discovery or theoretical breakthrough (deep neural networks and

had been described by many people, as far back as the 1950s) but because of two factors: the incredible increase in computer power (including the hundred-fold increase in speed by switching to

GPU A graphics processing unit (GPU) is a specialized electronic circuit designed to manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device. GPUs are used in embedded systems, mob ...

s) and the availability of vast amounts of training data, especially the giant curated datasets used for benchmark testing, such as ImageNet.

GPT

Generative pre-trained transformers (GPT) are

large language model A large language model (LLM) is a language model consisting of a neural network with many parameters (typically billions of weights or more), trained on large quantities of unlabelled text using self-supervised learning. LLMs emerged around 2018 an ...

s (LLMs) that generate text based on the semantic relationships between words in sentences. Text-based GPT models are pretrained on a large

corpus of text In linguistics, a corpus (plural ''corpora'') or text corpus is a language resource consisting of a large and structured set of texts (nowadays usually electronically stored and processed). In corpus linguistics, they are used to do statistical ...

that can be from the Internet. The pretraining consists of predicting the next

token Token may refer to: Arts, entertainment, and media * Token, a game piece or counter, used in some games * The Tokens, a vocal music group * Tolkien Black, a recurring character on the animated television series ''South Park,'' formerly known as ...

(a token being usually a word, subword, or punctuation). Throughout this pretraining, GPT models accumulate knowledge about the world and can then generate human-like text by repeatedly predicting the next token. Typically, a subsequent training phase makes the model more truthful, useful, and harmless, usually with a technique called

reinforcement learning from human feedback In machine learning, reinforcement learning from human feedback (RLHF) or reinforcement learning from human preferences is a technique that trains a "reward model" directly from human feedback and uses the model as a reward function to optimize an ...

(RLHF). Current GPT models are prone to generating falsehoods called "

hallucinations A hallucination is a perception in the absence of an external stimulus that has the qualities of a real perception. Hallucinations are vivid, substantial, and are perceived to be located in external objective space. Hallucination is a combinati ...

", although this can be reduced with RLHF and quality data. They are used in

chatbot A chatbot or chatterbot is a software application used to conduct an on-line chat conversation via text or text-to-speech, in lieu of providing direct contact with a live human agent. Designed to convincingly simulate the way a human would behav ...

s, which allow people to ask a question or request a task in simple text. Current models and services include

Gemini Gemini may refer to: Space * Gemini (constellation), one of the constellations of the zodiac ** Gemini in Chinese astronomy * Project Gemini, the second U.S. crewed spaceflight program * Gemini Observatory, consisting of telescopes in the Northern ...

(formerly Bard),

Grok ''Grok'' is a neologism coined by American writer Robert A. Heinlein for his 1961 science fiction novel ''Stranger in a Strange Land''. While the ''Oxford English Dictionary'' summarizes the meaning of ''grok'' as "to understand intuitively or ...

Claude Claude may refer to: __NOTOC__ People and fictional characters * Claude (given name), a list of people and fictional characters * Claude (surname), a list of people * Claude Lorrain (c. 1600–1682), French landscape painter, draughtsman and etcher ...

Copilot In aviation, the first officer (FO), also called co-pilot, is the pilot who is second-in-command of the aircraft to the captain, who is the legal commander. In the event of incapacitation of the captain, the first officer will assume command of ...

, and

LLaMA The llama (; ) (''Lama glama'') is a domesticated South American camelid, widely used as a meat and pack animal by Andean cultures since the Pre-Columbian era. Llamas are social animals and live with others as a herd. Their wool is so ...

. Multimodal GPT models can process different types of data ( modalities) such as images, videos, sound, and text.

Hardware and software

In the late 2010s,

graphics processing unit A graphics processing unit (GPU) is a specialized electronic circuit designed to manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device. GPUs are used in embedded systems, mo ...

s (GPUs) that were increasingly designed with AI-specific enhancements and used with specialized

TensorFlow TensorFlow is a free and open-source software library for machine learning and artificial intelligence. It can be used across a range of tasks but has a particular focus on training and inference of deep neural networks. "It is machine learning ...

software had replaced previously used

central processing unit A central processing unit (CPU), also called a central processor, main processor or just processor, is the electronic circuitry that executes instructions comprising a computer program. The CPU performs basic arithmetic, logic, controlling, an ...

(CPUs) as the dominant means for large-scale (commercial and academic)

machine learning Machine learning (ML) is a field of inquiry devoted to understanding and building methods that 'learn', that is, methods that leverage data to improve performance on some set of tasks. It is seen as a part of artificial intelligence. Machine ...

models' training. Specialized

programming language A programming language is a system of notation for writing computer programs. Most programming languages are text-based formal languages, but they may also be graphical. They are a kind of computer language. The description of a programming l ...

s such as

were used in early AI research, but

general-purpose programming language In computer software, a general-purpose programming language (GPL) is a programming language for building software in a wide variety of application domains. Conversely, a domain-specific programming language is used within a specific area. For ex ...

s like Python have become predominant. The transistor density in integrated circuits has been observed to roughly double every 18 months—a trend known as

Moore's law Moore's law is the observation that the number of transistors in a dense integrated circuit (IC) doubles about every two years. Moore's law is an observation and projection of a historical trend. Rather than a law of physics, it is an empi ...

, named after the

Intel Intel Corporation is an American multinational corporation and technology company headquartered in Santa Clara, California, Santa Clara, California. It is the world's largest semiconductor chip manufacturer by revenue, and is one of the devel ...

co-founder

Gordon Moore Gordon Earle Moore (born January 3, 1929) is an American businessman, engineer, and the co-founder and chairman emeritus of Intel Corporation. He is also the original proponent of Moore's law. As of March 2021, Moore's net worth is re ...

, who first identified it. Improvements in

GPUs A graphics processing unit (GPU) is a specialized electronic circuit designed to manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device. GPUs are used in embedded systems, mob ...

have been even faster.

Applications

AI and machine learning technology is used in most of the essential applications of the 2020s, including:

search engines A search engine is a software system designed to carry out web searches. They search the World Wide Web in a systematic way for particular information specified in a textual web search query. The search results are generally presented in a l ...

(such as

), targeting online advertisements,

(offered by

), driving

internet traffic Internet traffic is the flow of data within the entire Internet, or in certain network links of its constituent networks. Common traffic measurements are total volume, in units of multiples of the byte, or as transmission rates in bytes per cert ...

targeted advertising Targeted advertising is a form of advertising, including online advertising, that is directed towards an audience with certain traits, based on the product or person the advertiser is promoting. These traits can either be demographic with a focus ...

(

AdSense Google AdSense is a program run by Google through which website publishers in the Google Network of content sites serve text, images, video, or interactive media advertisements that are targeted to the site content and audience. These adver ...

Facebook Facebook is an online social media and social networking service owned by American company Meta Platforms. Founded in 2004 by Mark Zuckerberg with fellow Harvard College students and roommates Eduardo Saverin, Andrew McCollum, Dustin ...

s (such as

(including

drones Drone most commonly refers to: * Drone (bee), a male bee, from an unfertilized egg * Unmanned aerial vehicle * Unmanned surface vehicle, watercraft * Unmanned underwater vehicle or underwater drone Drone, drones or The Drones may also refer to: ...

, ADAS and

self-driving cars A self-driving car, also known as an autonomous car, driver-less car, or robotic car (robo-car), is a car that is capable of traveling without human input.Xie, S.; Hu, J.; Bhowmick, P.; Ding, Z.; Arvin, F.,Distributed Motion Planning for Sa ...

automatic language translation Machine translation, sometimes referred to by the abbreviation MT (not to be confused with computer-aided translation, machine-aided human translation or interactive translation), is a sub-field of computational linguistics that investigates th ...

(

Microsoft Translator Microsoft Translator is a multilingual machine translation cloud service provided by Microsoft. Microsoft Translator is a part of Microsoft Cognitive Services and integrated across multiple consumer, developer, and enterprise products; including ...

Google Translate Google Translate is a multilingual neural machine translation service developed by Google to translate text, documents and websites from one language into another. It offers a website interface, a mobile app for Android and iOS, and an A ...

(

Apple An apple is an edible fruit produced by an apple tree (''Malus domestica''). Apple trees are cultivated worldwide and are the most widely grown species in the genus '' Malus''. The tree originated in Central Asia, where its wild ances ...

Face ID Face ID is a facial recognition system designed and developed by Apple Inc. for the iPhone and iPad Pro. The system allows biometric authentication for unlocking a device, making payments, accessing sensitive data, providing detailed facial exp ...

Microsoft Microsoft Corporation is an American multinational corporation, multinational technology company, technology corporation producing Software, computer software, consumer electronics, personal computers, and related services headquartered at th ...

DeepFace DeepFace is a deep learning facial recognition system created by a research group at Facebook. It identifies human faces in digital images. The program employs a nine-layer neural network with over 120 million connection weights and was trained on ...

and

Google Google LLC () is an American Multinational corporation, multinational technology company focusing on Search Engine, search engine technology, online advertising, cloud computing, software, computer software, quantum computing, e-commerce, ar ...

FaceNet FaceNet is a facial recognition system developed by Florian Schroff, Dmitry Kalenichenko and James Philbina, a group of researchers affiliated with Google. The system was first presented at the 2015 IEEE Conference on Computer Vision and Pattern Re ...

) and

image labeling Automatic image annotation (also known as automatic image tagging or linguistic indexing) is the process by which a computer system automatically assigns metadata in the form of captioning or keywords to a digital image. This application of comput ...

(used by

, Apple's

iPhoto iPhoto is a discontinued digital photograph manipulation software application developed by Apple Inc. It was included with every Macintosh personal computer from 2002 to 2015, when it was replaced with Apple's Photos application. Originally s ...

and

TikTok TikTok, known in China as Douyin (), is a short-form video hosting service owned by the Chinese company ByteDance. It hosts user-submitted videos, which can range in duration from 15 seconds to 10 minutes. TikTok is an international version ...

). The deployment of AI may be overseen by a

Chief automation officer The chief automation officer (CAO) is the manager overseeing an enterprises's digital automation and robotic process automation strategy. The position came to prominence following the growth of artificial intelligence Artificial intellige ...

(CAO).

Health and medicine

The application of AI in

medicine Medicine is the science and Praxis (process), practice of caring for a patient, managing the diagnosis, prognosis, Preventive medicine, prevention, therapy, treatment, Palliative care, palliation of their injury or disease, and Health promotion ...

and

medical research Medical research (or biomedical research), also known as experimental medicine, encompasses a wide array of research, extending from " basic research" (also called ''bench science'' or ''bench research''), – involving fundamental scienti ...

has the potential to increase patient care and quality of life. Through the lens of the

Hippocratic Oath The Hippocratic Oath is an oath of ethics historically taken by physicians. It is one of the most widely known of Greek medical texts. In its original form, it requires a new physician to swear, by a number of healing gods, to uphold specific e ...

, medical professionals are ethically compelled to use AI, if applications can more accurately diagnose and treat patients. For medical research, AI is an important tool for processing and integrating big data. This is particularly important for

organoid An organoid is a miniaturized and simplified version of an organ produced in vitro in three dimensions that shows realistic micro-anatomy. They are derived from one or a few cells from a tissue, embryonic stem cells or induced pluripotent stem ...

and

tissue engineering Tissue engineering is a biomedical engineering discipline that uses a combination of cells, engineering, materials methods, and suitable biochemical and physicochemical factors to restore, maintain, improve, or replace different types of biologi ...

development which use

microscopy Microscopy is the technical field of using microscopes to view objects and areas of objects that cannot be seen with the naked eye (objects that are not within the resolution range of the normal eye). There are three well-known branches of mi ...

imaging as a key technique in fabrication. It has been suggested that AI can overcome discrepancies in funding allocated to different fields of research. New AI tools can deepen the understanding of biomedically relevant pathways. For example,

AlphaFold 2 AlphaFold is an artificial intelligence (AI) program developed by DeepMind, a subsidiary of Alphabet, which performs predictions of protein structure. The program is designed as a deep learning system. AlphaFold AI software has had two major v ...

(2021) demonstrated the ability to approximate, in hours rather than months, the 3D structure of a protein. In 2023, it was reported that AI-guided drug discovery helped find a class of antibiotics capable of killing two different types of drug-resistant bacteria. In 2024, researchers used machine learning to accelerate the search for

Parkinson's disease Parkinson's disease (PD), or simply Parkinson's, is a long-term degenerative disorder of the central nervous system that mainly affects the motor system. The symptoms usually emerge slowly, and as the disease worsens, non-motor symptoms becom ...

drug treatments. Their aim was to identify compounds that block the clumping, or aggregation, of

alpha-synuclein Alpha-synuclein is a protein that, in humans, is encoded by the ''SNCA'' gene. Alpha-synuclein is a neuronal protein that regulates synaptic vesicle trafficking and subsequent neurotransmitter release. It is abundant in the brain, while smaller a ...

(the protein that characterises Parkinson's disease). They were able to speed up the initial screening process ten-fold and reduce the cost by a thousand-fold.

Sexuality

Applications of AI in this domain include AI-enabled menstruation and fertility trackers that analyze user data to offer prediction, AI-integrated sex toys (e.g.,

teledildonics Teledildonics (also known as cyberdildonics) is the name coined for virtual sex encounters using networked, electronic sex toys to mimic and extend human sexual interaction. The term became known after technology critic and writer Howard Rheingol ...

), AI-generated sexual education content, and AI agents that simulate sexual and romantic partners (e.g., Replika). AI is also used for the production of non-consensual

deepfake pornography Deepfake pornography, or simply fake pornography, is a type of synthetic porn that is created via altering already-existing pornographic material by applying deepfake technology to the faces of the actor or actress. Deepfake porn has been very cont ...

, raising significant ethical and legal concerns. AI technologies have also been used to attempt to identify online gender-based violence and online

sexual grooming Sexual grooming refers to actions or behaviors used to establish an emotional connection with a minor, and sometimes the child's family, to lower the child's inhibitions with the objective of sexual abuse. It can occur in various settings, inclu ...

of minors.

Games

Game playing programs have been used since the 1950s to demonstrate and test AI's most advanced techniques.

Deep Blue Deep Blue may refer to: Film * '' Deep Blues: A Musical Pilgrimage to the Crossroads'', a 1992 documentary film about Mississippi Delta blues music * ''Deep Blue'' (2001 film), a film by Dwight H. Little * ''Deep Blue'' (2003 film), a film us ...

became the first computer chess-playing system to beat a reigning world chess champion,

Garry Kasparov Garry Kimovich Kasparov (born 13 April 1963) is a Russian chess Grandmaster (chess), grandmaster, former World Chess Champion, writer, political activist and commentator. His peak Elo rating system, rating of 2851, achieved in 1999, was the hi ...

, on 11 May 1997. In 2011, in a ''

Jeopardy! ''Jeopardy!'' is an American game show created by Merv Griffin. The show is a quiz competition that reverses the traditional question-and-answer format of many quiz shows. Rather than being given questions, contestants are instead given ge ...

quiz show A game show is a genre of broadcast viewing entertainment (radio, television, internet, stage or other) where contestants compete for a reward. These programs can either be participatory or demonstrative and are typically directed by a host, ...

exhibition match, IBM's

question answering system Question answering (QA) is a computer science discipline within the fields of information retrieval and natural language processing (NLP), which is concerned with building systems that automatically answer questions posed by humans in a natural l ...

Watson Watson may refer to: Companies * Actavis, a pharmaceutical company formerly known as Watson Pharmaceuticals * A.S. Watson Group, retail division of Hutchison Whampoa * Thomas J. Watson Research Center, IBM research center * Watson Systems, maker ...

, defeated the two greatest ''Jeopardy!'' champions,

Brad Rutter Bradford Gates Rutter (born January 31, 1978) is an American game show contestant, TV host, producer, and actor. With over $5.1 million in winnings, he is currently the 2nd highest-earning American game show contestant of all time, behind Ken ...

and

Ken Jennings Kenneth Wayne Jennings III (born May 23, 1974) is an American game show host, author, and former game show contestant. He is the highest-earning American game show contestant, having won money on five different game shows, including $4,522,70 ...

, by a significant margin. In March 2016,

AlphaGo AlphaGo is a computer program that plays the board game Go. It was developed by DeepMind Technologies a subsidiary of Google (now Alphabet Inc.). Subsequent versions of AlphaGo became increasingly powerful, including a version that competed u ...

won 4 out of 5 games of Go in a match with Go champion

Lee Sedol Lee Sedol ( ko, 이세돌; born 2 March 1983), or Lee Se-dol, is a former South Korean professional Go player of 9 dan rank. As of February 2016, he ranked second in international titles (18), behind only Lee Chang-ho (21). He is the f ...

, becoming the first

computer Go Computer Go is the field of artificial intelligence (AI) dedicated to creating a computer program that plays the traditional board game Go. The field is sharply divided into two eras. Before 2015, the programs of the era were weak. The best ...

-playing system to beat a professional Go player without handicaps. Then, in 2017, it defeated Ke Jie, who was the best Go player in the world. Other programs handle imperfect-information games, such as the

poker Poker is a family of comparing card games in which players wager over which hand is best according to that specific game's rules. It is played worldwide, however in some places the rules may vary. While the earliest known form of the game w ...

-playing program

Pluribus The Pluribus''Pluribus'' is the ablative plural of the Latin word for "more" or "above." multiprocessor was an early multi-processor computer designed by BBN for use as a packet switch in the ARPANET. Its design later influenced the BBN Butter ...

DeepMind DeepMind Technologies is a British artificial intelligence subsidiary of Alphabet Inc. and research laboratory founded in 2010. DeepMind was acquired by Google in 2014 and became a wholly owned subsidiary of Alphabet Inc, after Google's restru ...

developed increasingly generalistic

models, such as with

MuZero MuZero is a computer program developed by artificial intelligence research company DeepMind to master games without knowing their rules. Its release in 2019 included benchmarks of its performance in go, chess, shogi, and a standard suite of Ata ...

, which could be trained to play chess, Go, or Atari games. In 2019, DeepMind's AlphaStar achieved grandmaster level in

StarCraft II ''StarCraft II'' is a military science fiction video game created by Blizzard Entertainment as a sequel to the successful ''StarCraft'' video game released in 1998. Set in a fictional future, the game centers on a galactic struggle for dominance a ...

, a particularly challenging real-time strategy game that involves incomplete knowledge of what happens on the map. In 2021, an AI agent competed in a PlayStation Gran Turismo competition, winning against four of the world's best Gran Turismo drivers using deep reinforcement learning. In 2024, Google DeepMind introduced SIMA, a type of AI capable of autonomously playing nine previously unseen open-world video games by observing screen output, as well as executing short, specific tasks in response to natural language instructions.

Mathematics

In mathematics, special forms of formal step-by-step

are used. In contrast, LLMs such as ''

GPT-4 Generative Pre-trained Transformer 4 (GPT-4) is a multimodal large language model created by OpenAI and the fourth in its GPT series. It was released on March 14, 2023, and has been made publicly available in a limited form via ChatGPT Plus, ...

Turbo'', '' Gemini Ultra'', '' Claude Opus'', '' LLaMa-2'' or '' Mistral Large'' are working with probabilistic models, which can produce wrong answers in the form of

. Therefore, they need not only a large database of mathematical problems to learn from but also methods such as supervised

fine-tuning In theoretical physics, fine-tuning is the process in which parameters of a model must be adjusted very precisely in order to fit with certain observations. This had led to the discovery that the fundamental constants and quantities fall into suc ...

or trained classifiers with human-annotated data to improve answers for new problems and learn from corrections. A 2024 study showed that the performance of some language models for reasoning capabilities in solving math problems not included in their training data was low, even for problems with only minor deviations from trained data. Alternatively, dedicated models for mathematical problem solving with higher precision for the outcome including proof of theorems have been developed such as ''Alpha Tensor'', ''Alpha Geometry'' and ''Alpha Proof'' all from

Google DeepMind DeepMind Technologies is a British artificial intelligence subsidiary of Alphabet Inc. and research laboratory founded in 2010. DeepMind was acquired by Google in 2014 and became a wholly owned subsidiary of Alphabet Inc, after Google's restru ...

, ''Llemma'' from eleuther or ''Julius''. When natural language is used to describe mathematical problems, converters transform such prompts into a formal language such as

Lean Lean, leaning or LEAN may refer to: Business practices * Lean thinking, a business methodology adopted in various fields ** Lean construction, an adaption of lean manufacturing principles to the design and construction process ** Lean governme ...

to define mathematical tasks. Some models have been developed to solve challenging problems and reach good results in benchmark tests, others to serve as educational tools in mathematics.

Finance

Finance is one of the fastest growing sectors where applied AI tools are being deployed: from retail online banking to investment advice and insurance, where automated "robot advisers" have been in use for some years. World Pensions experts like Nicolas Firzli insist it may be too early to see the emergence of highly innovative AI-informed financial products and services: "the deployment of AI tools will simply further automatise things: destroying tens of thousands of jobs in banking, financial planning, and pension advice in the process, but I'm not sure it will unleash a new wave of .g., sophisticatedpension innovation."

Military

Various countries are deploying AI military applications. PD-notice The main applications enhance

command and control Command and control (abbr. C2) is a "set of organizational and technical attributes and processes ... hatemploys human, physical, and information resources to solve problems and accomplish missions" to achieve the goals of an organization or e ...

, communications, sensors, integration and interoperability. Research is targeting intelligence collection and analysis, logistics, cyber operations, information operations, and semiautonomous and

. AI technologies enable coordination of sensors and effectors, threat detection and identification, marking of enemy positions,

target acquisition Target acquisition is the detection and identification of the location of a target in sufficient detail to permit the effective employment of lethal and non-lethal means. The term is used for a broad area of applications. A "target" here is an e ...

, coordination and deconfliction of distributed Joint Fires between networked combat vehicles involving manned and unmanned teams. AI has been used in military operations in Iraq, Syria, Israel and Ukraine.

Generative AI

In the early 2020s,

generative AI Generative artificial intelligence (generative AI, GenAI, or GAI) is a subset of artificial intelligence that uses generative models to produce text, images, videos, or other forms of data. These models machine learning, learn the underlying p ...

gained widespread prominence. GenAI is AI capable of generating text, images, videos, or other data using

generative model In statistical classification, two main approaches are called the generative approach and the discriminative approach. These compute classifiers by different approaches, differing in the degree of statistical modelling. Terminology is incons ...

s, often in response to prompts. In March 2023, 58% of U.S. adults had heard about

and 14% had tried it. The increasing realism and ease-of-use of AI-based

text-to-image A text-to-image model is a machine learning model which takes as input a natural language description and produces an image matching that description. Such models began to be developed in the mid-2010s, as a result of advances in deep neural netwo ...

generators such as

Midjourney Midjourney is an independent research lab that produces an artificial intelligence program under the same name that creates images from textual descriptions, similar to OpenAI's DALL-E and Stable Diffusion. It is speculated that the underlying t ...

DALL-E DALL-E (stylized as DALL·E) and DALL-E 2 are deep learning models developed by OpenAI to generate digital images from natural language descriptions, called "prompts". DALL-E was revealed by OpenAI in a blog post in January 2021, and uses a ve ...

, and

Stable Diffusion Stable Diffusion is a deep learning, text-to-image model released in 2022. It is primarily used to generate detailed images conditioned on text descriptions, though it can also be applied to other tasks such as inpainting, outpainting, and genera ...

sparked a trend of

viral Viral means "relating to viruses" (small infectious agents). Viral may also refer to: Viral behavior, or virality Memetic behavior likened that of a virus, for example: * Viral marketing, the use of existing social networks to spread a marke ...

AI-generated photos. Widespread attention was gained by a fake photo of

Pope Francis Pope Francis ( la, Franciscus; it, Francesco; es, link=, Francisco; born Jorge Mario Bergoglio, 17 December 1936) is the head of the Catholic Church. He has been the bishop of Rome and sovereign of the Vatican City State since 13 March 2013. ...

wearing a white puffer coat, the fictional arrest of

Donald Trump Donald John Trump (born June 14, 1946) is an American politician, media personality, and businessman who served as the 45th president of the United States from 2017 to 2021. Trump graduated from the Wharton School of the University of ...

, and a hoax of an attack on the

Pentagon In geometry, a pentagon (from the Greek language, Greek πέντε ''pente'' meaning ''five'' and γωνία ''gonia'' meaning ''angle'') is any five-sided polygon or 5-gon. The sum of the internal angles in a simple polygon, simple pentagon is ...

, as well as the usage in professional creative arts.

Agents

Artificial intelligent (AI) agents are software entities designed to perceive their environment, make decisions, and take actions autonomously to achieve specific goals. These agents can interact with users, their environment, or other agents. AI agents are used in various applications, including

chatbots A chatbot or chatterbot is a software application used to conduct an on-line chat conversation via text or text-to-speech, in lieu of providing direct contact with a live human agent. Designed to convincingly simulate the way a human would behav ...

, game-playing systems, and

industrial robotics An industrial robot is a robot system used for manufacturing. Industrial robots are automated, programmable and capable of movement on three or more axes. Typical applications of robots include welding, painting, assembly, disassembly, pick ...

. AI agents operate within the constraints of their programming, available computational resources, and hardware limitations. This means they are restricted to performing tasks within their defined scope and have finite memory and processing capabilities. In real-world applications, AI agents often face time constraints for decision-making and action execution. Many AI agents incorporate learning algorithms, enabling them to improve their performance over time through experience or training. Using machine learning, AI agents can adapt to new situations and optimise their behaviour for their designated tasks.

Other industry-specific tasks

There are also thousands of successful AI applications used to solve specific problems for specific industries or institutions. In a 2017 survey, one in five companies reported having incorporated "AI" in some offerings or processes. A few examples are energy storage, medical diagnosis, military logistics, applications that predict the result of judicial decisions,

foreign policy A state's foreign policy or external policy (as opposed to internal or domestic policy) is its objectives and activities in relation to its interactions with other states, unions, and other political entities, whether bilaterally or through ...

, or supply chain management. AI applications for evacuation and

disaster A disaster is a serious problem occurring over a short or long period of time that causes widespread human, material, economic or environmental loss which exceeds the ability of the affected community or society to cope using its own resources ...

management are growing. AI has been used to investigate if and how people evacuated in large scale and small scale evacuations using historical data from GPS, videos or social media. Further, AI can provide real time information on the real time evacuation conditions. In agriculture, AI has helped farmers identify areas that need irrigation, fertilization, pesticide treatments or increasing yield. Agronomists use AI to conduct research and development. AI has been used to predict the ripening time for crops such as tomatoes, monitor soil moisture, operate agricultural robots, conduct

predictive analytics Predictive analytics encompasses a variety of statistical techniques from data mining, predictive modeling, and machine learning that analyze current and historical facts to make predictions about future or otherwise unknown events. In busin ...

, classify livestock pig call emotions, automate greenhouses, detect diseases and pests, and save water. Artificial intelligence is used in astronomy to analyze increasing amounts of available data and applications, mainly for "classification, regression, clustering, forecasting, generation, discovery, and the development of new scientific insights." For example, it is used for discovering exoplanets, forecasting solar activity, and distinguishing between signals and instrumental effects in gravitational wave astronomy. Additionally, it could be used for activities in space, such as space exploration, including the analysis of data from space missions, real-time science decisions of spacecraft, space debris avoidance, and more autonomous operation. During the 2024 Indian elections, US$50 millions was spent on authorized AI-generated content, notably by creating

deepfakes Deepfakes (a portmanteau of " deep learning" and "fake") are synthetic media in which a person in an existing image or video is replaced with someone else's likeness. While the act of creating fake content is not new, deepfakes leverage powerfu ...

of allied (including sometimes deceased) politicians to better engage with voters, and by translating speeches to various local languages.

Ethics

AI has potential benefits and potential risks. AI may be able to advance science and find solutions for serious problems:

Demis Hassabis Demis Hassabis (born 27 July 1976) is a British artificial intelligence researcher and entrepreneur. In his early career he was a video game AI programmer and designer, and an expert player of board games. He is the chief executive officer and ...

hopes to "solve intelligence, and then use that to solve everything else". However, as the use of AI has become widespread, several unintended consequences and risks have been identified. In-production systems can sometimes not factor ethics and bias into their AI training processes, especially when the AI algorithms are inherently unexplainable in deep learning.

Risks and harm

Privacy and copyright

Machine learning algorithms require large amounts of data. The techniques used to acquire this data have raised concerns about

privacy Privacy (, ) is the ability of an individual or group to seclude themselves or information about themselves, and thereby express themselves selectively. The domain of privacy partially overlaps with security, which can include the concepts of a ...

, surveillance and

copyright A copyright is a type of intellectual property that gives its owner the exclusive right to copy, distribute, adapt, display, and perform a creative work, usually for a limited time. The creative work may be in a literary, artistic, education ...

. AI-powered devices and services, such as virtual assistants and IoT products, continuously collect personal information, raising concerns about intrusive data gathering and unauthorized access by third parties. The loss of privacy is further exacerbated by AI's ability to process and combine vast amounts of data, potentially leading to a surveillance society where individual activities are constantly monitored and analyzed without adequate safeguards or transparency. Sensitive user data collected may include online activity records, geolocation data, video or audio. For example, in order to build

algorithms,

has recorded millions of private conversations and allowed

temporary worker Temporary work or temporary employment (also called gigs) refers to an employment situation where the working arrangement is limited to a certain period of time based on the needs of the employing organization. Temporary employees are sometimes ...

s to listen to and transcribe some of them. Opinions about this widespread surveillance range from those who see it as a necessary evil to those for whom it is clearly

unethical Ethics or moral philosophy is a branch of philosophy that "involves systematizing, defending, and recommending concepts of right and wrong behavior".''Internet Encyclopedia of Philosophy'' The field of ethics, along with aesthetics, concerns ...

and a violation of the

right to privacy The right to privacy is an element of various legal traditions that intends to restrain governmental and private actions that threaten the privacy Privacy (, ) is the ability of an individual or group to seclude themselves or information a ...

. AI developers argue that this is the only way to deliver valuable applications. and have developed several techniques that attempt to preserve privacy while still obtaining the data, such as

data aggregation Data aggregation is the compiling of information from databases with intent to prepare combined datasets for data processing. Description The United States Geological Survey explains that, “when data are well documented, you know how and where to ...

, de-identification and

differential privacy Differential privacy (DP) is a system for publicly sharing information about a dataset by describing the patterns of groups within the dataset while withholding information about individuals in the dataset. The idea behind differential privacy is t ...

. Since 2016, some privacy experts, such as

Cynthia Dwork Cynthia Dwork (born June 27, 1958) is an American computer scientist at Harvard University, where she is Gordon McKay Professor of Computer Science, Radcliffe Alumnae Professor at the Radcliffe Institute for Advanced Study, and Affiliated Professo ...

, have begun to view privacy in terms of

fairness Fairness or being fair can refer to: * Justice * The character in the award-nominated musical comedy '' A Theory of Justice: The Musical.'' * Equity (law), a legal principle allowing for the use of discretion and fairness when applying justice ...

Brian Christian Brian Christian (born 1984 in Wilmington, Delaware) is an American non-fiction author, poet, programmer and researcher, best known for a bestselling series of books about the human implications of computer science, including ''The Most Human Human ...

wrote that experts have pivoted "from the question of 'what they know' to the question of 'what they're doing with it'." Generative AI is often trained on unlicensed copyrighted works, including in domains such as images or computer code; the output is then used under the rationale of "

fair use Fair use is a doctrine in United States law that permits limited use of copyrighted material without having to first acquire permission from the copyright holder. Fair use is one of the limitations to copyright intended to balance the intere ...

". Experts disagree about how well and under what circumstances this rationale will hold up in courts of law; relevant factors may include "the purpose and character of the use of the copyrighted work" and "the effect upon the potential market for the copyrighted work". Website owners who do not wish to have their content scraped can indicate it in a "

robots.txt The robots exclusion standard, also known as the robots exclusion protocol or simply robots.txt, is a standard used by websites to indicate to visiting web crawlers and other web robots which portions of the site they are allowed to visit. Th ...

" file. In 2023, leading authors (including

John Grisham John Ray Grisham Jr. (; born February 8, 1955 in Jonesboro, Arkansas) is an American novelist, lawyer and former member of the 7th district of the Mississippi House of Representatives, known for his popular legal thrillers. According to the Am ...

and

Jonathan Franzen Jonathan Earl Franzen (born August 17, 1959) is an American novelist and essayist. His 2001 novel '' The Corrections'', a sprawling, satirical family drama, drew widespread critical acclaim, earned Franzen a National Book Award, was a Pulitzer Pr ...

) sued AI companies for using their work to train generative AI. Another discussed approach is to envision a separate '' sui generis'' system of protection for creations generated by AI to ensure fair attribution and compensation for human authors.

Dominance by tech giants

The commercial AI scene is dominated by

Big Tech Big Tech, also known as the Tech Giants, refers to the most dominant companies in the information technology industry, mostly located in the United States. The term also refers to the four or five largest American tech companies, called the B ...

companies such as

Alphabet Inc. Alphabet Inc. is an American multinational technology conglomerate holding company headquartered in Mountain View, California. It was created through a restructuring of Google on October 2, 2015, and became the parent company of Google and sev ...

Apple Inc. Apple Inc. is an American multinational technology company headquartered in Cupertino, California, United States. Apple is the largest technology company by revenue (totaling in 2021) and, as of June 2022, is the world's biggest company ...

Meta Platforms Meta Platforms, Inc., (file no. 3835815) doing business as Meta and formerly named Facebook, Inc., and TheFacebook, Inc., is an American multinational technology conglomerate based in Menlo Park, California. The company owns Facebook, Instag ...

, and

. Some of these players already own the vast majority of existing

cloud infrastructure Cloud computing is the on-demand availability of computer system resources, especially data storage (cloud storage) and computing power, without direct active management by the user. Large clouds often have functions distributed over mul ...

and

computing Computing is any goal-oriented activity requiring, benefiting from, or creating computing machinery. It includes the study and experimentation of algorithmic processes, and development of both hardware and software. Computing has scientific, ...

power from data centers, allowing them to entrench further in the marketplace.

Power needs and environmental impacts

In January 2024, the

International Energy Agency The International Energy Agency (IEA) is a Paris-based autonomous Intergovernmental organization, intergovernmental organisation, established in 1974, that provides policy recommendations, analysis and data on the entire global energy sector, wit ...

(IEA) released ''Electricity 2024, Analysis and Forecast to 2026'', forecasting electric power use. This is the first IEA report to make projections for data centers and power consumption for artificial intelligence and cryptocurrency. The report states that power demand for these uses might double by 2026, with additional electric power usage equal to electricity used by the whole Japanese nation. Prodigious power consumption by AI is responsible for the growth of fossil fuels use, and might delay closings of obsolete, carbon-emitting coal energy facilities. There is a feverish rise in the construction of data centers throughout the US, making large technology firms (e.g., Microsoft, Meta, Google, Amazon) into voracious consumers of electric power. Projected electric consumption is so immense that there is concern that it will be fulfilled no matter the source. A ChatGPT search involves the use of 10 times the electrical energy as a Google search. The large firms are in haste to find power sources – from nuclear energy to geothermal to fusion. The tech firms argue that – in the long view – AI will be eventually kinder to the environment, but they need the energy now. AI makes the power grid more efficient and "intelligent", will assist in the growth of nuclear power, and track overall carbon emissions, according to technology firms. A 2024 Goldman Sachs Research Paper, ''AI Data Centers and the Coming US Power Demand Surge'', found "US power demand (is) likely to experience growth not seen in a generation...." and forecasts that, by 2030, US data centers will consume 8% of US power, as opposed to 3% in 2022, presaging growth for the electrical power generation industry by a variety of means. Data centers' need for more and more electrical power is such that they might max out the electrical grid. The Big Tech companies counter that AI can be used to maximize the utilization of the grid by all. In 2024, the ''Wall Street Journal'' reported that big AI companies have begun negotiations with the US nuclear power providers to provide electricity to the data centers. In March 2024 Amazon purchased a Pennsylvania nuclear-powered data center for $650 Million (US).

Nvidia Nvidia CorporationOfficially written as NVIDIA and stylized in its logo as VIDIA with the lowercase "n" the same height as the uppercase "VIDIA"; formerly stylized as VIDIA with a large italicized lowercase "n" on products from the mid 1990s to ...

CEO

Jen-Hsun Huang Jen-Hsun "Jensen" Huang (; born February 17, 1963) is a Taiwanese American billionaire business magnate, electrical engineer, and the co-founder, current president and CEO of Nvidia Corporation. Early years and education Huang was born in ...

said nuclear power is a good option for the data centers. In September 2024,

announced an agreement with

Constellation Energy Constellation Energy Corporation () is an energy company headquartered in Baltimore, Maryland, United States. The company provides electric power, natural gas, and energy management services. It has approximately two million customers across th ...

to re-open the

Three Mile Island 3 is a number, numeral, and glyph. 3, three, or III may also refer to: * AD 3, the third year of the AD era * 3 BC, the third year before the AD era * March, the third month Books * '' Three of Them'' (Russian: ', literally, "three"), a 1901 ...

nuclear power plant to provide Microsoft with 100% of all electric power produced by the plant for 20 years. Reopening the plant, which suffered a partial nuclear meltdown of its Unit 2 reactor in 1979, will require Constellation to get through strict regulatory processes which will include extensive safety scrutiny from the US

Nuclear Regulatory Commission The Nuclear Regulatory Commission (NRC) is an independent agency of the United States government tasked with protecting public health and safety related to nuclear energy. Established by the Energy Reorganization Act of 1974, the NRC began oper ...

. If approved (this will be the first ever US re-commissioning of a nuclear plant), over 835 megawatts of power – enough for 800,000 homes – of energy will be produced. The cost for re-opening and upgrading is estimated at $1.6 billion (US) and is dependent on tax breaks for nuclear power contained in the 2022 US

Inflation Reduction Act The Inflation Reduction Act of 2022 (IRA) is a landmark United States federal law which aims to curb inflation by reducing the deficit, lowering prescription drug prices, and investing into domestic energy production while promoting clean ener ...

. The US government and the state of Michigan are investing almost $2 billion (US) to reopen the Palisades Nuclear reactor on Lake Michigan. Closed since 2022, the plant is planned to be reopened in October 2025. The Three Mile Island facility will be renamed the Crane Clean Energy Center after Chris Crane, a nuclear proponent and former CEO of

Exelon Exelon Corporation is an American Fortune 100 energy company headquartered in Chicago, Illinois and incorporated in Pennsylvania. It generates revenues of approximately $33.5 billion and employs approximately 33,400 people. Exelon is the larges ...

who was responsible for Exelon spinoff of Constellation. After the last approval in September 2023,

Taiwan Taiwan, officially the Republic of China (ROC), is a country in East Asia, at the junction of the East and South China Seas in the northwestern Pacific Ocean, with the People's Republic of China (PRC) to the northwest, Japan to the northe ...

suspended the approval of data centers north of Taoyuan with a capacity of more than 5 MW in 2024, due to power supply shortages. Taiwan aims to phase out nuclear power by 2025. On the other hand,

Singapore Singapore (), officially the Republic of Singapore, is a sovereign island country and city-state in maritime Southeast Asia. It lies about one degree of latitude () north of the equator, off the southern tip of the Malay Peninsula, borde ...

imposed a ban on the opening of data centers in 2019 due to electric power, but in 2022, lifted this ban. Although most nuclear plants in Japan have been shut down after the 2011

Fukushima nuclear accident The was a nuclear accident in 2011 at the Fukushima Daiichi Nuclear Power Plant in Ōkuma, Fukushima, Japan. The proximate cause of the disaster was the 2011 Tōhoku earthquake and tsunami, which occurred on the afternoon of 11 March 2011 and ...

, according to an October 2024 ''Bloomberg'' article in Japanese, cloud gaming services company Ubitus, in which Nvidia has a stake, is looking for land in Japan near nuclear power plant for a new data center for generative AI. Ubitus CEO Wesley Kuo said nuclear power plants are the most efficient, cheap and stable power for AI. On 1 November 2024, the

Federal Energy Regulatory Commission The Federal Energy Regulatory Commission (FERC) is the United States federal agency that regulates the transmission and wholesale sale of electricity and natural gas in interstate commerce and regulates the transportation of oil by pipeline in ...

(FERC) rejected an application submitted by

Talen Energy Transcription activator-like effector nucleases (TALEN) are restriction enzymes that can be engineered to cut specific sequences of DNA. They are made by fusing a TAL effector DNA-binding domain to a DNA cleavage domain (a nuclease which cuts ...

for approval to supply some electricity from the nuclear power station Susquehanna to Amazon's data center. According to the Commission Chairman

Willie L. Phillips Willie L. Phillips is an American attorney who served as the chairman of the Federal Energy Regulatory Commission (FERC) from 2023 to 2025. Career Phillips is originally from Fairhope, Alabama, and attended the University of Montevallo. Phil ...

, it is a burden on the electricity grid as well as a significant cost shifting concern to households and other business sectors.

Misinformation

and others use

recommender system A recommender system, or a recommendation system (sometimes replacing 'system' with a synonym such as platform or engine), is a subclass of information filtering system that provide suggestions for items that are most pertinent to a particular ...

s to guide users to more content. These AI programs were given the goal of

maximizing Maximization is a style of decision-making characterized by seeking the best option through an exhaustive search through alternatives. It is contrasted with satisficing, in which individuals evaluate options until they find one that is "good enough" ...

user engagement (that is, the only goal was to keep people watching). The AI learned that users tended to choose

misinformation Misinformation is incorrect or misleading information. It differs from disinformation, which is ''deliberately'' deceptive. Rumors are information not attributed to any particular source, and so are unreliable and often unverified, but can turn ou ...

conspiracy theories A conspiracy theory is an explanation for an event or situation that invokes a conspiracy by sinister and powerful groups, often political in motivation, when other explanations are more probable.Additional sources: * * * * The term has a neg ...

, and extreme

partisan Partisan may refer to: Military * Partisan (weapon), a pole weapon * Partisan (military), paramilitary forces engaged behind the front line Films * ''Partisan'' (film), a 2015 Australian film * '' Hell River'', a 1974 Yugoslavian film also kno ...

content, and, to keep them watching, the AI recommended more of it. Users also tended to watch more content on the same subject, so the AI led people into

filter bubbles A filter bubble or ideological frame is a state of intellectual isolationTechnopediaDefinition – What does Filter Bubble mean?, Retrieved October 10, 2017, "....A filter bubble is the intellectual isolation, that can occur when websites make us ...

where they received multiple versions of the same misinformation. This convinced many users that the misinformation was true, and ultimately undermined trust in institutions, the media and the government. The AI program had correctly learned to maximize its goal, but the result was harmful to society. After the U.S. election in 2016, major technology companies took steps to mitigate the problem . In 2022,

began to create images, audio, video and text that are indistinguishable from real photographs, recordings, films, or human writing. It is possible for bad actors to use this technology to create massive amounts of misinformation or propaganda. AI pioneer

Geoffrey Hinton Geoffrey Everest Hinton One or more of the preceding sentences incorporates text from the royalsociety.org website where: (born 6 December 1947) is a British-Canadian cognitive psychologist and computer scientist, most noted for his work on ...

expressed concern about AI enabling "authoritarian leaders to manipulate their electorates" on a large scale, among other risks.

Algorithmic bias and fairness

Machine learning applications will be biased if they learn from biased data. The developers may not be aware that the bias exists. Bias can be introduced by the way

training data In machine learning, a common task is the study and construction of algorithms that can learn from and make predictions on data. Such algorithms function by making data-driven predictions or decisions, through building a mathematical model from ...

is selected and by the way a model is deployed. If a biased algorithm is used to make decisions that can seriously

harm Harm is a moral and legal concept. Bernard Gert construes harm as any of the following: * pain * death * disability * mortality * loss of abil ity or freedom * loss of pleasure. Joel Feinberg gives an account of harm as setbacks to in ...

people (as it can in

, finance,

recruitment Recruitment is the overall process of identifying, sourcing, screening, shortlisting, and interviewing candidates for jobs (either permanent or temporary) within an organization. Recruitment also is the processes involved in choosing individua ...

housing Housing, or more generally, living spaces, refers to the construction and housing authority, assigned usage of houses or buildings individually or collectively, for the purpose of Shelter (building), shelter. Housing ensures that members of so ...

policing The police are a constituted body of persons empowered by a state, with the aim to enforce the law, to ensure the safety, health and possessions of citizens, and to prevent crime and civil disorder. Their lawful powers include arrest and t ...

) then the algorithm may cause discrimination. The field of

studies how to prevent harms from algorithmic biases. On June 28, 2015,

Google Photos Google Photos is a photo sharing and storage service developed by Google. It was announced in May 2015 and spun off from Google+, the company's former social network. As of June 1, 2021, in its free tier, any newly uploaded photo and video c ...

's new image labeling feature mistakenly identified Jacky Alcine and a friend as "gorillas" because they were black. The system was trained on a dataset that contained very few images of black people, a problem called "sample size disparity". Google "fixed" this problem by preventing the system from labelling ''anything'' as a "gorilla". Eight years later, in 2023, Google Photos still could not identify a gorilla, and neither could similar products from Apple, Facebook, Microsoft and Amazon.

COMPAS Compas, also known as compas direct or compas direk (; Haitian Creole: ''konpa'', ''kompa'' or ''kompa dirèk''), is a modern méringue dance music genre of Haiti. The genre was popularized following the creation of Ensemble Aux Callebasses i ...

is a commercial program widely used by

U.S. court The courts of the United States are closely linked hierarchical systems of courts at the federal and state levels. The federal courts form the judicial branch of the US government and operate under the authority of the United States Constitution an ...

s to assess the likelihood of a

defendant In court proceedings, a defendant is a person or object who is the party either accused of committing a crime in criminal prosecution or against whom some type of civil relief is being sought in a civil case. Terminology varies from one juris ...

becoming a

recidivist Recidivism (; from ''recidive'' and ''ism'', from Latin ''recidīvus'' "recurring", from ''re-'' "back" and ''cadō'' "I fall") is the act of a person repeating an undesirable behavior after they have experienced negative consequences of th ...

. In 2016,

Julia Angwin Julia Angwin is a Pulitzer Prize-winning American investigative journalist, New York Times bestselling author, and entrepreneur. She is co-founder and editor-in-chief of The Markup, a nonprofit newsroom that investigates the impact of technology ...

ProPublica ProPublica (), legally Pro Publica, Inc., is a nonprofit organization based in New York City. In 2010, it became the first online news source to win a Pulitzer Prize, for a piece written by one of its journalists''The Guardian'', April 13, 2010 ...

discovered that COMPAS exhibited racial bias, despite the fact that the program was not told the races of the defendants. Although the error rate for both whites and blacks was calibrated equal at exactly 61%, the errors for each race were different—the system consistently overestimated the chance that a black person would re-offend and would underestimate the chance that a white person would not re-offend. In 2017, several researchers showed that it was mathematically impossible for COMPAS to accommodate all possible measures of fairness when the base rates of re-offense were different for whites and blacks in the data. A program can make biased decisions even if the data does not explicitly mention a problematic feature (such as "race" or "gender"). The feature will correlate with other features (like "address", "shopping history" or "first name"), and the program will make the same decisions based on these features as it would on "race" or "gender". Moritz Hardt said "the most robust fact in this research area is that fairness through blindness doesn't work." Criticism of COMPAS highlighted that machine learning models are designed to make "predictions" that are only valid if we assume that the future will resemble the past. If they are trained on data that includes the results of racist decisions in the past, machine learning models must predict that racist decisions will be made in the future. If an application then uses these predictions as ''recommendations'', some of these "recommendations" will likely be racist. Thus, machine learning is not well suited to help make decisions in areas where there is hope that the future will be ''better'' than the past. It is descriptive rather than prescriptive. Bias and unfairness may go undetected because the developers are overwhelmingly white and male: among AI engineers, about 4% are black and 20% are women. There are various conflicting definitions and mathematical models of fairness. These notions depend on ethical assumptions, and are influenced by beliefs about society. One broad category is distributive fairness, which focuses on the outcomes, often identifying groups and seeking to compensate for statistical disparities. Representational fairness tries to ensure that AI systems do not reinforce negative

stereotype In social psychology, a stereotype is a generalized belief about a particular category of people. It is an expectation that people might have about every person of a particular group. The type of expectation can vary; it can be, for exampl ...

s or render certain groups invisible. Procedural fairness focuses on the decision process rather than the outcome. The most relevant notions of fairness may depend on the context, notably the type of AI application and the stakeholders. The subjectivity in the notions of bias and fairness makes it difficult for companies to operationalize them. Having access to sensitive attributes such as race or gender is also considered by many AI ethicists to be necessary in order to compensate for biases, but it may conflict with

anti-discrimination law Anti-discrimination law or non-discrimination law refers to legislation designed to prevent discrimination against particular groups of people; these groups are often referred to as protected groups or protected classes. Anti-discrimination laws ...

s. At its 2022 Conference on Fairness, Accountability, and Transparency (ACM FAccT 2022), the

Association for Computing Machinery The Association for Computing Machinery (ACM) is a US-based international learned society for computing. It was founded in 1947 and is the world's largest scientific and educational computing society. The ACM is a non-profit professional member ...

, in Seoul, South Korea, presented and published findings that recommend that until AI and robotics systems are demonstrated to be free of bias mistakes, they are unsafe, and the use of self-learning neural networks trained on vast, unregulated sources of flawed internet data should be curtailed.

Lack of transparency

Many AI systems are so complex that their designers cannot explain how they reach their decisions. Particularly with

deep neural networks Deep learning (also known as deep structured learning) is part of a broader family of machine learning methods based on artificial neural networks with representation learning. Learning can be supervised, semi-supervised or unsupervised. ...

, in which there are a large amount of non-

linear Linearity is the property of a mathematical relationship ('' function'') that can be graphically represented as a straight line. Linearity is closely related to '' proportionality''. Examples in physics include rectilinear motion, the linear ...

relationships between inputs and outputs. But some popular explainability techniques exist. It is impossible to be certain that a program is operating correctly if no one knows how exactly it works. There have been many cases where a machine learning program passed rigorous tests, but nevertheless learned something different than what the programmers intended. For example, a system that could identify skin diseases better than medical professionals was found to actually have a strong tendency to classify images with a

ruler A ruler, sometimes called a rule, line gauge, or scale, is a device used in geometry and technical drawing, as well as the engineering and construction industries, to measure distances or draw straight lines. Variants Rulers have long ...

as "cancerous", because pictures of malignancies typically include a ruler to show the scale. Another machine learning system designed to help effectively allocate medical resources was found to classify patients with asthma as being at "low risk" of dying from pneumonia. Having asthma is actually a severe risk factor, but since the patients having asthma would usually get much more medical care, they were relatively unlikely to die according to the training data. The correlation between asthma and low risk of dying from pneumonia was real, but misleading. People who have been harmed by an algorithm's decision have a right to an explanation. Doctors, for example, are expected to clearly and completely explain to their colleagues the reasoning behind any decision they make. Early drafts of the European Union's General Data Protection Regulation in 2016 included an explicit statement that this right exists. Industry experts noted that this is an unsolved problem with no solution in sight. Regulators argued that nevertheless the harm is real: if the problem has no solution, the tools should not be used.

DARPA The Defense Advanced Research Projects Agency (DARPA) is a research and development agency of the United States Department of Defense responsible for the development of emerging technologies for use by the military. Originally known as the Ad ...

established the XAI ("Explainable Artificial Intelligence") program in 2014 to try to solve these problems. Several approaches aim to address the transparency problem. SHAP enables to visualise the contribution of each feature to the output. LIME can locally approximate a model's outputs with a simpler, interpretable model.

Multitask learning Multi-task learning (MTL) is a subfield of machine learning in which multiple learning tasks are solved at the same time, while exploiting commonalities and differences across tasks. This can result in improved learning efficiency and prediction ac ...

provides a large number of outputs in addition to the target classification. These other outputs can help developers deduce what the network has learned.

Deconvolution In mathematics, deconvolution is the operation inverse to convolution. Both operations are used in signal processing and image processing. For example, it may be possible to recover the original signal after a filter (convolution) by using a de ...

DeepDream DeepDream is a computer vision program created by Google engineer Alexander Mordvintsev that uses a convolutional neural network to find and enhance patterns in images via algorithmic pareidolia, thus creating a dream-like appearance reminiscent ...

and other

methods can allow developers to see what different layers of a deep network for computer vision have learned, and produce output that can suggest what the network is learning. For generative pre-trained transformers,

Anthropic {{Short pages monitor Artificial intelligence, Artificial intelligence Computational fields of study Computational neuroscience Cybernetics Data science Formal sciences Intelligence by type