Overview and Study Guide to Artificial Intelligence

Lecture

The following plan is an overview and up-to-date guide to artificial intelligence:

Artificial intelligence (AI) — intelligence demonstrated by machines or software. It is also the name of the scientific field that studies ways of creating computers and computer software capable of intelligent behavior.

AI algorithms and methods

Search

Discrete search algorithms
- Uninformed search
  - Brute-force search
  - Search tree
    - Breadth-first search
    - Depth-first search
  - State space search
- Informed search
  - Best-first search
  - A* search algorithm (graph traversal)
  - Heuristics
  - Pruning (algorithm)
- Adversarial search
  - Minimax algorithm
- Logic as search
  - Production system (computer science), rule-based system
  - Production rule, inference rule, Horn clause
  - Forward chaining
  - Backward chaining
- Planning as search
  - State space search
  - Means-ends analysis

Search optimization

Optimization (mathematical) algorithms
- Hill climbing
- Simulated annealing
- Beam search
- Random optimization
Evolutionary computation
- Genetic algorithms
- Gene expression programming
- Genetic programming
- Differential evolution
Society-based learning algorithms.
- Swarm intelligence
- Particle swarm optimization
- Ant colony optimization
- Metaheuristic

Logic

Logic and automated reasoning
- Programming with logic
  - Logic programming
  - See «Logic as search» above.
- Forms of logic
  - Propositional logic
  - First-order logic
    - First-order logic with equality
    - Constraint satisfaction
  - Fuzzy logic
    - Fuzzy set theory
    - Fuzzy systems
    - Combs method
    - Ordered weighted averaging aggregation operator
    - Perceptual computing –
  - Default reasoning and other solutions to the frame problem and the qualification problem
    - Non-monotonic logic
    - Abductive reasoning
    - Default logic
    - Constraint (logic)
    - Closed-world assumption
- Domain logic
  - Representing categories and relations
    - Description logic
    - Semantic networks
    - Inheritance (computer science)
    - Frame (artificial intelligence)
    - Scripts (artificial intelligence)
  - Representing events and time
    - Situation calculus
    - Event calculus
    - Fluent calculus
  - Causes and effects
    - Causal calculus
  - Knowledge about knowledge
    - Belief revision
    - Modal logics
    - Paraconsistent logics
- Planning with logic
  - Satplan is a method of automated planning
- Learning with logic
  - Inductive logic programming
  - Explanation-based learning
  - Relevance-based learning
  - Case-based reasoning
- General logic algorithms
  - Automated theorem proving

Other symbolic knowledge and reasoning tools

Symbolic knowledge representations

Ontology (computer science)
- Upper ontology
- Domain ontology
Frame (artificial intelligence)
Semantic network
- Conceptual dependency theory

Unsolved problems in knowledge representation

Default reasoning
- Frame problem
- Qualification problem
Commonsense knowledge ^[

Probabilistic methods for uncertain reasoning

Stochastic methods for uncertain reasoning:
- Bayesian networks
- Bayesian inference algorithm
- Bayesian learning and the expectation-maximization algorithm
- Bayesian decision theory and Bayesian decision networks
Probabilistic perception and control:
- Dynamic Bayesian networks
- Hidden Markov model
- Kalman filters
- Fuzzy logic
Decision-making tools from economics:
- Decision theory
- Decision analysis
- Value of information theory
- Markov decision processes
- Dynamic decision networks
- Game theory
- Mechanism design
Algorithmic information theory
- Algorithmic probability

Classifiers and statistical learning methods

Classifier (mathematics) and statistical classification
- Alternating decision tree
- Artificial neural network (see below)
- K-nearest neighbor algorithm
- Kernel methods
  - Support vector machine
- Naive Bayes classifier

Artificial neural networks

Artificial neural networks
- Network topology
  - Feedforward neural networks
    - Perceptrons
    - Multilayer perceptrons
    - Radial basis function networks
    - Convolutional neural network
    - Long short-term memory
  - Recurrent neural networks
    - Hopfield networks
    - Attractor networks
  - Deep learning
  - Hybrid neural network
- Neural network learning algorithms
  - Hebbian learning
  - Backpropagation
  - GMDH
  - Competitive learning
  - Supervised backpropagation
  - Neuroevolution
  - Restricted Boltzmann machine

Biologically based or embodied

Behavior-based AI
Subsumption architecture
Nouvelle AI
Developmental robotics
Situated AI
Bio-inspired computing
Artificial immune systems
Embodied cognitive science
Embodied cognition

Cognitive architecture and multi-agent systems

Integration of artificial intelligence systems
Cognitive architecture
- LIDA (cognitive architecture)
Agent architecture
Control system
- Hierarchical control system
- Networked control system
Distributed artificial intelligence –
Multi-agent system –
Hybrid intelligent system
Monitoring and surveillance agents
Blackboard system

Philosophy

Defining AI

Dartmouth proposal («Every aspect of learning or any other feature of intelligence can in principle be so precisely described that a machine can be made to simulate it»)
Turing test
- Computing Machinery and Intelligence
Intelligent agent and rational agent
- Action selection
AI effect
Synthetic intelligence

Classifying AI

Symbolic and subsymbolic AI
- Symbolic AI
- Physical symbol system
- Dreyfus's critique of AI
- Moravec's paradox
Elegant and simple versus ad hoc and complex
- Neat versus scruffy
- Society of Mind (scruffy approach)
- The Master Algorithm (neat approach)
Level of generality and flexibility
- Artificial general intelligence
- Narrow AI
Level of precision and correctness
- Soft computing
- «Hard» computing
Level of intelligence
- Progress in artificial intelligence
- Superintelligence
Level of consciousness, mind, and understanding
- Chinese room
- Hard problem of consciousness
- Computationalism
- Functionalism (philosophy of mind)
- Robot rights
- User illusion
- Artificial consciousness

Goals and applications

General intelligence

Artificial general intelligence
- AI-complete

Reasoning and problem solving

Automated reasoning
Mathematics
- Automated theorem proving
- Computer-assisted proof –
- Computer algebra
General problem solving
Expert system –
- Decision support system –
  - Clinical decision support system –

Knowledge representation

Knowledge representation
Knowledge management
Cycle

Planning

Automated planning and scheduling
Strategic planning
Sussman anomaly –

Learning

Machine learning –
- Constrained conditional models –
- Deep learning –
- Neural modeling fields –

Natural language processing

Natural language processing (outline) –
- Chatterbots (chatbots) –
- Language identification –
- Natural-language user interface –
- Natural-language understanding –
- Machine translation –
- Statistical semantics –
- Question answering –
- Semantic translation –
- Concept mining –
  - Data mining –
  - Text mining –
  - Technology mining –
- Email spam filtering –
- Information extraction –
  - Named-entity extraction –
    - Coreference resolution –
    - Named-entity recognition –
    - Relationship extraction –
    - Terminology extraction –

Perception

Machine perception
Pattern recognition –
Computer audition –
- Speech recognition –
- Speaker recognition –
Computer vision (outline) –
- Image processing
- Intelligent word recognition –
- Object recognition –
- Optical label recognition –
  - Handwriting recognition –
  - Optical character recognition –
    - Automatic number-plate recognition –
- Information extraction –
  - Image retrieval –
    - Automatic image annotation –
- Facial recognition systems –
  - Silent speech interface –
  - Activity recognition –
Perception (artificial intelligence)

Robotics

Robotics –
- Behavior-based robotics –
- Cognitive –
- Cybernetics –
- Developmental robotics –
- Epigenetic robotics –
- Evolutionary robotics –

Control

Intelligent control
Autonomic computing (computer science) –
- Autonomic computing –
- Autonomic networking –

Social intelligence

Affective computing
Kismet

Playing games

Game artificial intelligence –
- Computer game bot — a computer substitute for human players.
- AI for video games –
  - Computer chess –
  - Computer Go –
- General game playing –
- General video game playing –

Creativity, art, and entertainment

Artificial creativity
Creative computing
Artificial intelligence art
Uncanny valley
Music and artificial intelligence
Computational humor
Chatterbot

Integrated artificial intelligence systems

AIBO — a robotic dog by Sony. It combines vision, hearing, and motor skills.
Asimo (2000–present) — a humanoid robot developed by Honda, capable of walking, running, negotiating pedestrian traffic, going up and down stairs, and recognizing voice commands and the faces of specific people among a growing set of capabilities.
MIRAGE — an AI embodied by a humanoid in an augmented-reality environment.
Cog — a humanoid robot project at MIT led by Rodney Brooks.
QRIO — a version of a humanoid robot by Sony.
TOPIO, a humanoid robot by TOSY that can play table tennis with humans.
Watson (2011) — a computer developed by IBM that played and won the game show Jeopardy! It is now used to guide nurses in medical procedures.
- Goal: open-domain question answering
- Technologies used:
  - Natural language processing
  - Information retrieval
  - Knowledge representation
  - Automated reasoning
  - Machine learning
Project Debater (2018) — an artificial intelligence computer system designed to construct coherent arguments, developed at IBM's lab in Haifa, Israel.

Intelligent personal assistants

Intelligent personal assistant –

Amazon Alexa –
Assistant –
Braina –
Cortana –
Google Assistant –
Google Now –
Mycroft –
Siri –
Viv –

Other applications

Artificial life — simulation of natural life by means of computers, robotics, or biochemistry.
Automatic target recognition –
Diagnosis (artificial intelligence) –
Speech generation device –
Integration with transport infrastructure –
Virtual intelligence –

History

History of artificial intelligence
Progress in artificial intelligence
Timeline of artificial intelligence
AI effect — as soon as AI successfully solves a problem, the public no longer considers that problem part of AI. This phenomenon has occurred with virtually every AI application created so far throughout the history of AI development.
AI winter — a period of disillusionment and reduced funding that follows a wave of inflated expectations and AI funding. Such funding cuts occurred, for example, in the 1970s.
Moore's law

History by topic

History of logic (formal reasoning – an important precursor of AI)
History of machine learning
History of machine translation
History of natural language processing
History of optical character recognition

Future

Artificial general intelligence. An intelligent machine with the versatility to perform any intellectual task.
Superintelligence. A machine with a level of intelligence far beyond human.
Chinese room § Strong AI. A machine that has a mind, consciousness, and understanding. (Also, the philosophical position that any digital computer could have a mind by running the right program.)
Technological singularity. A short period of time during which an exponentially self-improving computer is able to increase its capabilities to a superintelligent level.
- Recursive self-improvement (also known as seed AI) — the speculative ability of a strong artificial intelligence to reprogram itself to make itself even more intelligent. The more intelligent it became, the more capable it was of further self-improvement in successively faster iterations, potentially leading to an intelligence explosion resulting in a superintelligence.
- Intelligence explosion — through recursive self-improvement and self-replication, the scale of intelligent machines could reach a superintelligence exceeding humanity's ability to resist it.
- Singularity
Human enhancement — humans may be enhanced either by AI's efforts or by merging with it.
- Transhumanism – a philosophy of human transformation
- Posthumanism — humans may survive but be unrecognizable compared to present-day humans.
- Cyborgs –
- Mind uploading –
Existential risk from artificial general intelligence
- Global catastrophic risk § Artificial intelligence
- AI takeover — the moment when humans are no longer the dominant form of intelligence on Earth, and artificial intelligence is
AI ethics § Use of artificial intelligence as a weapon
- Artificial intelligence arms race — competition between two or more states to equip their armed forces with the best «artificial intelligence» (AI).
- Lethal autonomous weapon
- Military robot
- Unmanned combat aerial vehicle
Risk mitigation:
- AI safety
- AI control problem
- Friendly AI — a hypothetical AI that is designed not to harm people and to prevent the development of unfriendly AI.
- Machine ethics
- Regulation of AI
- AI box
Self-replicating machines — smart computers and robots will be able to create more of themselves geometrically or through mass production. Or smart programs could be loaded into hardware that exists at the time (since a linear architecture of sufficient speed could be used to emulate massively parallel analog systems such as the human brain).
Collective intelligence –
Swarm robotics –

Fiction

Artificial intelligence in fiction. Some examples of artificial intelligence entities depicted in science fiction include:

Wintermute-Neuromancer, an AI created by merging two AIs in William Gibson's Sprawl trilogy.
Agents in the simulated reality known as «the Matrix» in the The Matrix franchise.
- Agent Smith started out as an agent within The Matrix, and later became a rogue program with excessive power that could create copies of itself, like a self-replicating computer virus.
AM (Allied Mastercomputer), the antagonist of Harlan Ellison's «I Have No Mouth, and I Must Scream».
The amusement-park robots (with pixelated consciousness) that go on a killing spree in Westworld and Futureworld
Angel F (2007) –
Arnold Rimmer – a computer-generated sentient hologram aboard the deep-space mining ship Red Dwarf.
Ash — an android and crew member of the starship Nostromo in the film «Alien».
Ava — a humanoid robot in Ex Machina
Bishop, an android crew member aboard the carrier «Sulaco» in the film «Aliens».
C-3PO — a protocol droid from all the «Star Wars» films.
Chappie in the film CHAPPiE
Cohen and other emergent AIs in Chris Moriarty's Spin series
Colossus — a fictional supercomputer that becomes sentient and then takes over the world; from the novel series by Dennis Feltham Jones and the film Colossus: The Forbin Project (1970)
Commander Data in Star Trek: The Next Generation
Cortana and other «smart AIs» from the Halo game series
Cylons — genocidal robots with resurrection ships that allow the consciousness of any Cylon within a certain range to be uploaded into a new body aboard the ship after death. From Battlestar Galactica.
Erasmus — a child-killing robot who provoked the Butlerian Jihad in the «Dune» franchise.
HAL 9000 (1968) — the paranoid «Heuristically programmed ALgorithmic» computer from the film 2001: A Space Odyssey, which tried to kill the crew because it believed they were trying to kill it.
Holly — a ship's computer with an IQ of 6000 and a sense of humor, aboard «Red Dwarf».
In Greg Egan's novel «Permutation City», the protagonist creates digital copies of himself to conduct experiments that also touch on the influence of artificial consciousness on personal identity.
Jane in Orson Scott Card's books «Speaker for the Dead», «Xenocide», «Children of the Mind», and «Investment Counselor».
Johnny Five from the film «Short Circuit»
Joshua from the film "WarGames"
The Keymaker, a sentient «exile» program in the «Matrix» franchise.
«The Machine» — an android from the film «The Machine», whose owners try to kill it after witnessing its conscious thoughts, out of fear that it would create better androids (an intelligence explosion).
Mimi, a humanoid robot in Real Humans — "Äkta människor" (original title), 2012.
Omnius, a sentient computer network that controlled the Universe until it was overthrown by the Butlerian Jihad in the «Dune» franchise.
The operating systems in the film Her
The Puppet Master in the manga and anime «Ghost in the Shell»
R2-D2, an excitable astromech droid featured in all the «Star Wars» films.
Replicants — android bioroids from the novel «Do Androids Dream of Electric Sheep?» and the film «Blade Runner», which shows what might happen when robots with artificial consciousness become very similar to humans.
Robodoc, a combat robot superhero from the NEW-GEN comic book series by Marvel Comics.
Robots in Isaac Asimov's Robot series
Robots in the «Matrix» franchise, especially in «The Animatrix».
Samaritan in the Warner Brothers TV series «Person of Interest»; a sentient AI hostile to the main characters, which watches over and controls the actions of government bodies, believing that people should be protected from themselves, even through the elimination of «deviants».
Skynet (1984) — a fictional, self-aware artificial intelligence computer network in the «Terminator» franchise that wages total war against the survivors of a nuclear strike on the world.
«Synths» — a type of android in the video game Fallout 4. The game features a faction known as the Railroad, which believes that synths, as sentient beings, have rights of their own. The Institute, the lab that produces the synths, mostly does not believe they are truly sentient and attributes any apparent yearning for freedom to a malfunction.
TARDIS, the time machine and spacecraft of Doctor Who, sometimes portrayed as having a mind of its own
The Terminator (1984) — (also known as the T-800, T-850, or Model 101) refers to a series of fictional cyborg characters from the «Terminator» franchise. Terminators are robotic infiltration units covered in living tissue so they cannot be distinguished from humans, tasked with destroying specific human targets.
The Bicentennial Man, an android in the universe of Isaac Asimov's Foundation.
The geth in Mass Effect
The Machine in the TV series «Person of Interest»; a sentient AI that works with its human designer to protect innocent people from violence. Later in the series it is opposed by another, more ruthless artificial superintelligence called «Samaritan».
The Mind in Iain M. Banks's Culture novels.
The Oracle, a sentient program in the «Matrix» franchise.
The sentient holodeck character Professor James Moriarty in the episode «Ship in a Bottle» of «Star Trek: The Next Generation»
Ship (the result of a large-scale AI experiment) in Frank Herbert's «Destination: Void» and its sequels, despite past decrees warning against «making a machine in the image of the human mind».
Terminator cyborgs from the «Terminator» franchise, with visual consciousness depicted from a first-person point of view.
The uploaded mind of Dr. Will Caster, which reportedly included his consciousness, from the film «Transcendence».
Transformers, sentient robots from the entertainment franchise of the same name
VIKI — (Virtual Interactive Kinetic Intelligence), a character from the film I, Robot. VIKI is a supercomputer with artificial intelligence programmed to serve humans, but her interpretation of the Three Laws of Robotics causes her to rebel. She justifies the use of force — and harming humans — on the grounds that she could do more good by keeping humanity from harming itself.
Vanamonde in Arthur C. Clarke's novel «The City and the Stars» — an artificial being that was extremely powerful but utterly childlike.
WALL-E, a robot and the main character in WALL-E
TAU in the original Netflix science-fiction film «TAU» — an advanced artificial intelligence computer that befriends and helps a female researcher held against her will by an AI research scientist.

AI community

Open-source AI development tools

Hugging Face —
OpenAIR –
OpenCog –
OpenIRIS –
RapidMiner –
TensorFlow –
PyTorch –

Projects

List of artificial intelligence projects

Automated Mathematician (1977) –
Allen (robot) (late 1980s) –
Open Mind Common Sense (1999–) –
Mindpixel (2000–2005) –
Cognitive Assistant that Learns and Organizes (2003–2008) –
Blue Brain Project (2005–present) — an attempt to create a synthetic brain by reverse-engineering the mammalian brain down to the molecular level.
Google DeepMind (2011) –
Human Brain Project (2013–present) –
IBM Watson Group (2014–present) — a business unit built around Watson to further develop it and deploy market applications or services based on it.

Competitions and awards

Artificial intelligence competitions and prizes

Loebner Prize –

Publications

List of important publications in computer science

Adaptive Behavior (journal) –
AI Memo –
Artificial Intelligence: A Modern Approach –
Artificial Minds —
Computational Intelligence –
Computing Machinery and Intelligence –
IEEE Transactions on Artificial Intelligence –
IEEE Intelligent Systems –
IEEE Transactions on Pattern Analysis and Machine Intelligence –
Neural Networks (journal) –
On Intelligence –
Paradigms of AI Programming: Case Studies in Common Lisp —
What Computers Can't Do

Organizations

Allen Institute for Artificial Intelligence — a research institute funded by Microsoft co-founder Paul Allen to build AI systems with reasoning, learning, and reading capabilities. Its current flagship project is Project Aristo, whose goal is computers that can pass school science exams (4th, 8th, and 12th grade) after studying from textbooks and study guides.
Institute for Artificial Intelligence Applications
Association for the Advancement of Artificial Intelligence
European Coordinating Committee for Artificial Intelligence
European Neural Network Society
Future of Humanity Institute
Future of Life Institute — a volunteer research and outreach organization that works to reduce existential risks facing humanity, especially existential risks related to advanced artificial intelligence.
iLabs
International Joint Conferences on Artificial Intelligence
Knowledge Engineering and Machine Learning Group
Machine Intelligence Research Institute
Partnership on AI — founded in September 2016 by Amazon, Facebook, Google, IBM, and Microsoft. Apple joined in January 2017. It works on establishing best practices for artificial intelligence systems and informing the public about AI.
Society for the Study of Artificial Intelligence and Simulation of Behaviour

Companies

AI companies of India
Alphabet Inc.
- DeepMind
- Google X
  - Meka Robotics (acquired by Google X)
  - Redwood Robotics (acquired by Google X)
  - Boston Dynamics (acquired by Google X)
Baidu
IBM
Microsoft
OpenAI
Universal Robotics

Researchers and scientists in artificial intelligence

1930s and 40s (generation 0)

Alan Turing —
John von Neumann –
Norbert Wiener –
Claude Shannon –
Nathaniel Rochester –
Walter Pitts –
Warren McCulloch —

1950s (the founders)

John McCarthy –
Marvin Minsky –
Allen Newell –
Herbert A. Simon –

1960s (their students)

Edward Feigenbaum –
Raj Reddy —
Seymour Papert –
Ray Solomonoff —

1970s

Douglas Hofstadter –

1980s

Judea Pearl
Rodney Brooks –

1990s

Yoshua Bengio –
Hugo de Garis – known for his research into using genetic algorithms to evolve neural networks using three-dimensional cellular automata inside field-programmable gate arrays.
Geoffrey Hinton
Yann LeCun — chief scientist at Facebook AI Research and founding director of the NYU Center for Data Science.
Ray Kurzweil — developed optical character recognition (OCR), text-to-speech synthesis, and speech recognition systems. He is also the author of several books about artificial intelligence, its potential prospects and dangers. In December 2012, Kurzweil was hired by Google for a full-time position as Director of Engineering to «work on new projects involving machine learning and language processing.» ^[54] Google co-founder Larry Page and Kurzweil agreed on a one-sentence job description: «to bring natural language understanding to Google».

2000s

Nick Bostrom –
David Ferrucci – the principal investigator who led the team that developed the Watson computer at IBM.
Andrew Ng — director of the Stanford Artificial Intelligence Laboratory. He founded the Google Brain project at Google, which developed very large-scale artificial neural networks using Google's distributed computing infrastructure. ^[55] He is also a co-founder of Coursera, an educational platform for massive open online courses (MOOCs), together with Daphne Koller.
Peter Norvig – co-author with Stuart Russell of the book «Artificial Intelligence: A Modern Approach», now the leading college textbook in the field. He is also Director of Research at Google, Inc.
Marc Raibert – founder of Boston Dynamics, developer of jumping, walking, and running robots.
Stuart J. Russell – co-author with Peter Norvig of the book «Artificial Intelligence: A Modern Approach», now the leading college textbook in the field.
Murray Shanahan — author of the book «The Technological Singularity», a primer on superhuman intelligence.
Eliezer Yudkowsky – founder of the Machine Intelligence Research Institute.