Daily Archives: June 3, 2017

Date Development Antiquity Greek myths of Hephaestus and Pygmalion incorporated the idea of intelligent robots (such as Talos) and artificial beings (such as Galatea and Pandora).[1] Antiquity Yan Shi presented King Mu of Zhou with mechanical men.[2] Antiquity Sacred mechanical statues built in Egypt and Greece were believed to be capable of wisdom and emotion. Hermes Trismegistus would write "they have sensus and spiritus ... by discovering the true nature of the gods, man has been able to reproduce it." Mosaic law prohibits the use of automatons in religion.[3] 384 BC322 BC Aristotle described the syllogism, a method of formal, mechanical thought. 1st century Heron of Alexandria created mechanical men and other automatons.[4] 260 Porphyry of Tyros wrote Isagog which categorized knowledge and logic.[5] ~800 Geber develops the Arabic alchemical theory of Takwin, the artificial creation of life in the laboratory, up to and including human life.[6] 1206 Al-Jazari created a programmable orchestra of mechanical human beings.[7] 1275 Ramon Llull, Spanish theologian invents the Ars Magna, a tool for combining concepts mechanically, based on an Arabic astrological tool, the Zairja. The method would be developed further by Gottfried Leibniz in the 17th century.[8] ~1500 Paracelsus claimed to have created an artificial man out of magnetism, sperm and alchemy.[9] ~1580 Rabbi Judah Loew ben Bezalel of Prague is said to have invented the Golem, a clay man brought to life.[10] Early 17th century Ren Descartes proposed that bodies of animals are nothing more than complex machines (but that mental phenomena are of a different "substance").[11] 1623 Wilhelm Schickard drew a calculating clock on a letter to Kepler. This will be the first of five unsuccessful attempts at designing a direct entry calculating clock in the 17th century (including the designs of Tito Burattini, Samuel Morland and Ren Grillet)).[12] 1641 Thomas Hobbes published Leviathan and presented a mechanical, combinatorial theory of cognition. He wrote "...for reason is nothing but reckoning".[13][14] 1642 Blaise Pascal invented the mechanical calculator,[15] the first digital calculating machine[16] 1672 Gottfried Leibniz improved the earlier machines, making the Stepped Reckoner to do multiplication and division. He also invented the binary numeral system and envisioned a universal calculus of reasoning (alphabet of human thought) by which arguments could be decided mechanically. Leibniz worked on assigning a specific number to each and every object in the world, as a prelude to an algebraic solution to all possible problems.[17] 1726 Jonathan Swift published Gulliver's Travels, which includes this description of the Engine, a machine on the island of Laputa: "a Project for improving speculative Knowledge by practical and mechanical Operations " by using this "Contrivance", "the most ignorant Person at a reasonable Charge, and with a little bodily Labour, may write Books in Philosophy, Poetry, Politicks, Law, Mathematicks, and Theology, with the least Assistance from Genius or study."[18] The machine is a parody of Ars Magna, one of the inspirations of Gottfried Leibniz' mechanism. 1750 Julien Offray de La Mettrie published L'Homme Machine, which argued that human thought is strictly mechanical.[19] 1769 Wolfgang von Kempelen built and toured with his chess-playing automaton, The Turk.[20] The Turk was later shown to be a hoax, involving a human chess player. 1818 Mary Shelley published the story of Frankenstein; or the Modern Prometheus, a fictional consideration of the ethics of creating sentient beings.[21] 18221859 Charles Babbage & Ada Lovelace worked on programmable mechanical calculating machines.[22] 1837 The mathematician Bernard Bolzano made the first modern attempt to formalize semantics. 1854 George Boole set out to "investigate the fundamental laws of those operations of the mind by which reasoning is performed, to give expression to them in the symbolic language of a calculus", inventing Boolean algebra.[23] 1863 Samuel Butler suggested that Darwinian evolution also applies to machines, and speculates that they will one day become conscious and eventually supplant humanity.[24] Date Development 1913 Bertrand Russell and Alfred North Whitehead published Principia Mathematica, which revolutionized formal logic. 1915 Leonardo Torres y Quevedo built a chess automaton, El Ajedrecista and published speculation about thinking and automata.[25] 1923 Karel apek's play R.U.R. (Rossum's Universal Robots) opened in London. This is the first use of the word "robot" in English.[26] 1920s and 1930s Ludwig Wittgenstein and Rudolf Carnap lead philosophy into logical analysis of knowledge. Alonzo Church develops Lambda Calculus to investigate computability using recursive functional notation. 1931 Kurt Gdel showed that sufficiently powerful formal systems, if consistent, permit the formulation of true theorems that are unprovable by any theorem-proving machine deriving all possible theorems from the axioms. To do this he had to build a universal, integer-based programming language, which is the reason why he is sometimes called the "father of theoretical computer science". 1941 Konrad Zuse built the first working program-controlled computers.[27] 1943 Warren Sturgis McCulloch and Walter Pitts publish "A Logical Calculus of the Ideas Immanent in Nervous Activity" (1943), laying foundations for artificial neural networks.[28] 1943 Arturo Rosenblueth, Norbert Wiener and Julian Bigelow coin the term "cybernetics". Wiener's popular book by that name published in 1948. 1945 Game theory which would prove invaluable in the progress of AI was introduced with the 1944 paper, Theory of Games and Economic Behavior by mathematician John von Neumann and economist Oskar Morgenstern. 1945 Vannevar Bush published As We May Think (The Atlantic Monthly, July 1945) a prescient vision of the future in which computers assist humans in many activities. 1948 John von Neumann (quoted by E.T. Jaynes) in response to a comment at a lecture that it was impossible for a machine to think: "You insist that there is something a machine cannot do. If you will tell me precisely what it is that a machine cannot do, then I can always make a machine which will do just that!". Von Neumann was presumably alluding to the Church-Turing thesis which states that any effective procedure can be simulated by a (generalized) computer. Date Development 1950 Alan Turing proposes the Turing Test as a measure of machine intelligence.[29] 1950 Claude Shannon published a detailed analysis of chess playing as search. 1950 Isaac Asimov published his Three Laws of Robotics. 1951 The first working AI programs were written in 1951 to run on the Ferranti Mark 1 machine of the University of Manchester: a checkers-playing program written by Christopher Strachey and a chess-playing program written by Dietrich Prinz. 19521962 Arthur Samuel (IBM) wrote the first game-playing program,[30] for checkers (draughts), to achieve sufficient skill to challenge a respectable amateur. His first checkers-playing program was written in 1952, and in 1955 he created a version that learned to play.[31] 1956 The first Dartmouth College summer AI conference is organized by John McCarthy, Marvin Minsky, Nathan Rochester of IBM and Claude Shannon. 1956 The name artificial intelligence is used for the first time as the topic of the second Dartmouth Conference, organized by John McCarthy[32] 1956 The first demonstration of the Logic Theorist (LT) written by Allen Newell, J.C. Shaw and Herbert A. Simon (Carnegie Institute of Technology, now [[Carnegie Mellon University] or CMU]). This is often called the first AI program, though Samuel's checkers program also has a strong claim. 1957 The General Problem Solver (GPS) demonstrated by Newell, Shaw and Simon while at CMU. 1958 John McCarthy (Massachusetts Institute of Technology or MIT) invented the Lisp programming language. 1958 Herbert Gelernter and Nathan Rochester (IBM) described a theorem prover in geometry that exploits a semantic model of the domain in the form of diagrams of "typical" cases. 1958 Teddington Conference on the Mechanization of Thought Processes was held in the UK and among the papers presented were John McCarthy's Programs with Common Sense, Oliver Selfridge's Pandemonium, and Marvin Minsky's Some Methods of Heuristic Programming and Artificial Intelligence. 1959 John McCarthy and Marvin Minsky founded the MIT AI Lab. Late 1950s, early 1960s Margaret Masterman and colleagues at University of Cambridge design semantic nets for machine translation. Date Development 1960s Ray Solomonoff lays the foundations of a mathematical theory of AI, introducing universal Bayesian methods for inductive inference and prediction. 1960 Man-Computer Symbiosis by J.C.R. Licklider. 1961 James Slagle (PhD dissertation, MIT) wrote (in Lisp) the first symbolic integration program, SAINT, which solved calculus problems at the college freshman level. 1961 In Minds, Machines and Gdel, John Lucas[33] denied the possibility of machine intelligence on logical or philosophical grounds. He referred to Kurt Gdel's result of 1931: sufficiently powerful formal systems are either inconsistent or allow for formulating true theorems unprovable by any theorem-proving AI deriving all provable theorems from the axioms. Since humans are able to "see" the truth of such theorems, machines were deemed inferior. 1961 Unimation's industrial robot Unimate worked on a General Motors automobile assembly line. 1963 Thomas Evans' program, ANALOGY, written as part of his PhD work at MIT, demonstrated that computers can solve the same analogy problems as are given on IQ tests. 1963 Edward Feigenbaum and Julian Feldman published Computers and Thought, the first collection of articles about artificial intelligence. 1963 Leonard Uhr and Charles Vossler published "A Pattern Recognition Program That Generates, Evaluates, and Adjusts Its Own Operators", which described one of the first machine learning programs that could adaptively acquire and modify features and thereby overcome the limitations of simple perceptrons of Rosenblatt 1964 Danny Bobrow's dissertation at MIT (technical report #1 from MIT's AI group, Project MAC), shows that computers can understand natural language well enough to solve algebra word problems correctly. 1964 Bertram Raphael's MIT dissertation on the SIR program demonstrates the power of a logical representation of knowledge for question-answering systems. 1965 J. Alan Robinson invented a mechanical proof procedure, the Resolution Method, which allowed programs to work efficiently with formal logic as a representation language. 1965 Joseph Weizenbaum (MIT) built ELIZA, an interactive program that carries on a dialogue in English language on any topic. It was a popular toy at AI centers on the ARPANET when a version that "simulated" the dialogue of a psychotherapist was programmed. 1965 Edward Feigenbaum initiated Dendral, a ten-year effort to develop software to deduce the molecular structure of organic compounds using scientific instrument data. It was the first expert system. 1966 Ross Quillian (PhD dissertation, Carnegie Inst. of Technology, now CMU) demonstrated semantic nets. 1966 Machine Intelligence workshop at Edinburgh the first of an influential annual series organized by Donald Michie and others. 1966 Negative report on machine translation kills much work in Natural language processing (NLP) for many years. 1967 Dendral program (Edward Feigenbaum, Joshua Lederberg, Bruce Buchanan, Georgia Sutherland at Stanford University) demonstrated to interpret mass spectra on organic chemical compounds. First successful knowledge-based program for scientific reasoning. 1968 Joel Moses (PhD work at MIT) demonstrated the power of symbolic reasoning for integration problems in the Macsyma program. First successful knowledge-based program in mathematics. 1968 Richard Greenblatt (programmer) at MIT built a knowledge-based chess-playing program, MacHack, that was good enough to achieve a class-C rating in tournament play. 1968 Wallace and Boulton's program, Snob (Comp.J. 11(2) 1968), for unsupervised classification (clustering) uses the Bayesian Minimum Message Length criterion, a mathematical realisation of Occam's razor. 1969 Stanford Research Institute (SRI): Shakey the Robot, demonstrated combining animal locomotion, perception and problem solving. 1969 Roger Schank (Stanford) defined conceptual dependency model for natural language understanding. Later developed (in PhD dissertations at Yale University) for use in story understanding by Robert Wilensky and Wendy Lehnert, and for use in understanding memory by Janet Kolodner. 1969 Yorick Wilks (Stanford) developed the semantic coherence view of language called Preference Semantics, embodied in the first semantics-driven machine translation program, and the basis of many PhD dissertations since such as Bran Boguraev and David Carter at Cambridge. 1969 First International Joint Conference on Artificial Intelligence (IJCAI) held at Stanford. 1969 Marvin Minsky and Seymour Papert publish Perceptrons, demonstrating previously unrecognized limits of this feed-forward two-layered structure. This book is considered by some to mark the beginning of the AI winter of the 1970s, a failure of confidence and funding for AI. Nevertheless, significant progress in the field continued (see below). 1969 McCarthy and Hayes started the discussion about the frame problem with their essay, "Some Philosophical Problems from the Standpoint of Artificial Intelligence". Date Development Early 1970s Jane Robinson and Don Walker established an influential Natural Language Processing group at SRI. 1970 Jaime Carbonell (Sr.) developed SCHOLAR, an interactive program for computer assisted instruction based on semantic nets as the representation of knowledge. 1970 Bill Woods described Augmented Transition Networks (ATN's) as a representation for natural language understanding. 1970 Patrick Winston's PhD program, ARCH, at MIT learned concepts from examples in the world of children's blocks. 1971 Terry Winograd's PhD thesis (MIT) demonstrated the ability of computers to understand English sentences in a restricted world of children's blocks, in a coupling of his language understanding program, SHRDLU, with a robot arm that carried out instructions typed in English. 1971 Work on the Boyer-Moore theorem prover started in Edinburgh.[34] 1972 Prolog programming language developed by Alain Colmerauer. 1972 Earl Sacerdoti developed one of the first hierarchical planning programs, ABSTRIPS. 1973 The Assembly Robotics Group at University of Edinburgh builds Freddy Robot, capable of using visual perception to locate and assemble models. (See Edinburgh Freddy Assembly Robot: a versatile computer-controlled assembly system.) 1973 The Lighthill report gives a largely negative verdict on AI research in Great Britain and forms the basis for the decision by the British government to discontinue support for AI research in all but two universities. 1974 Ted Shortliffe's PhD dissertation on the MYCIN program (Stanford) demonstrated a very practical rule-based approach to medical diagnoses, even in the presence of uncertainty. While it borrowed from DENDRAL, its own contributions strongly influenced the future of expert system development, especially commercial systems. 1975 Earl Sacerdoti developed techniques of partial-order planning in his NOAH system, replacing the previous paradigm of search among state space descriptions. NOAH was applied at SRI International to interactively diagnose and repair electromechanical systems. 1975 Austin Tate developed the Nonlin hierarchical planning system able to search a space of partial plans characterised as alternative approaches to the underlying goal structure of the plan. 1975 Marvin Minsky published his widely read and influential article on Frames as a representation of knowledge, in which many ideas about schemas and semantic links are brought together. 1975 The Meta-Dendral learning program produced new results in chemistry (some rules of mass spectrometry) the first scientific discoveries by a computer to be published in a refereed journal. Mid-1970s Barbara Grosz (SRI) established limits to traditional AI approaches to discourse modeling. Subsequent work by Grosz, Bonnie Webber and Candace Sidner developed the notion of "centering", used in establishing focus of discourse and anaphoric references in Natural language processing. Mid-1970s David Marr and MIT colleagues describe the "primal sketch" and its role in visual perception. 1976 Douglas Lenat's AM program (Stanford PhD dissertation) demonstrated the discovery model (loosely guided search for interesting conjectures). 1976 Randall Davis demonstrated the power of meta-level reasoning in his PhD dissertation at Stanford. 1978 Tom Mitchell, at Stanford, invented the concept of Version spaces for describing the search space of a concept formation program. 1978 Herbert A. Simon wins the Nobel Prize in Economics for his theory of bounded rationality, one of the cornerstones of AI known as "satisficing". 1978 The MOLGEN program, written at Stanford by Mark Stefik and Peter Friedland, demonstrated that an object-oriented programming representation of knowledge can be used to plan gene-cloning experiments. 1979 Bill VanMelle's PhD dissertation at Stanford demonstrated the generality of MYCIN's representation of knowledge and style of reasoning in his EMYCIN program, the model for many commercial expert system "shells". 1979 Jack Myers and Harry Pople at University of Pittsburgh developed INTERNIST, a knowledge-based medical diagnosis program based on Dr. Myers' clinical knowledge. 1979 Cordell Green, David Barstow, Elaine Kant and others at Stanford demonstrated the CHI system for automatic programming. 1979 The Stanford Cart, built by Hans Moravec, becomes the first computer-controlled, autonomous vehicle when it successfully traverses a chair-filled room and circumnavigates the Stanford AI Lab. 1979 BKG, a backgammon program written by Hans Berliner at CMU, defeats the reigning world champion. 1979 Drew McDermott and Jon Doyle at MIT, and John McCarthy at Stanford begin publishing work on non-monotonic logics and formal aspects of truth maintenance. Late 1970s Stanford's SUMEX-AIM resource, headed by Ed Feigenbaum and Joshua Lederberg, demonstrates the power of the ARPAnet for scientific collaboration. Date Development 1980s Lisp machines developed and marketed. First expert system shells and commercial applications. 1980 First National Conference of the American Association for Artificial Intelligence (AAAI) held at Stanford. 1981 Danny Hillis designs the connection machine, which utilizes Parallel computing to bring new power to AI, and to computation in general. (Later founds Thinking Machines Corporation) 1982 The Fifth Generation Computer Systems project (FGCS), an initiative by Japan's Ministry of International Trade and Industry, begun in 1982, to create a "fifth generation computer" (see history of computing hardware) which was supposed to perform much calculation utilizing massive parallelism. 1983 John Laird and Paul Rosenbloom, working with Allen Newell, complete CMU dissertations on Soar (program). 1983 James F. Allen invents the Interval Calculus, the first widely used formalization of temporal events. Mid-1980s Neural Networks become widely used with the Backpropagation algorithm (first described by Paul Werbos in 1974). 1985 The autonomous drawing program, AARON, created by Harold Cohen, is demonstrated at the AAAI National Conference (based on more than a decade of work, and with subsequent work showing major developments). 1986 The team of Ernst Dickmanns at Bundeswehr University of Munich builds the first robot cars, driving up to 55mph on empty streets. 1986 Barbara Grosz and Candace Sidner create the first computation model of discourse, establishing the field of research.[35] 1987 Marvin Minsky published The Society of Mind, a theoretical description of the mind as a collection of cooperating agents. He had been lecturing on the idea for years before the book came out (c.f. Doyle 1983).[36] 1987 Around the same time, Rodney Brooks introduced the subsumption architecture and behavior-based robotics as a more minimalist modular model of natural intelligence; Nouvelle AI. 1987 Commercial launch of generation 2.0 of Alacrity by Alacritous Inc./Allstar Advice Inc. Toronto, the first commercial strategic and managerial advisory system. The system was based upon a forward-chaining, self-developed expert system with 3,000 rules about the evolution of markets and competitive strategies and co-authored by Alistair Davidson and Mary Chung, founders of the firm with the underlying engine developed by Paul Tarvydas. The Alacrity system also included a small financial expert system that interpreted financial statements and models.[37] 1989 Dean Pomerleau at CMU creates ALVINN (An Autonomous Land Vehicle in a Neural Network). Date Development Early 1990s TD-Gammon, a backgammon program written by Gerry Tesauro, demonstrates that reinforcement (learning) is powerful enough to create a championship-level game-playing program by competing favorably with world-class players. 1990s Major advances in all areas of AI, with significant demonstrations in machine learning, intelligent tutoring, case-based reasoning, multi-agent planning, scheduling, uncertain reasoning, data mining, natural language understanding and translation, vision, virtual reality, games, and other topics. 1991 DART scheduling application deployed in the first Gulf War paid back DARPA's investment of 30 years in AI research.[38] 1993 Ian Horswill extended behavior-based robotics by creating Polly, the first robot to navigate using vision and operate at animal-like speeds (1 meter/second). 1993 Rodney Brooks, Lynn Andrea Stein and Cynthia Breazeal started the widely publicized MIT Cog project with numerous collaborators, in an attempt to build a humanoid robot child in just five years. 1993 ISX corporation wins "DARPA contractor of the year"[39] for the Dynamic Analysis and Replanning Tool (DART) which reportedly repaid the US government's entire investment in AI research since the 1950s.[40] 1994 With passengers on board, the twin robot cars VaMP and VITA-2 of Ernst Dickmanns and Daimler-Benz drive more than one thousand kilometers on a Paris three-lane highway in standard heavy traffic at speeds up to 130km/h. They demonstrate autonomous driving in free lanes, convoy driving, and lane changes left and right with autonomous passing of other cars. 1994 English draughts (checkers) world champion Tinsley resigned a match against computer program Chinook. Chinook defeated 2nd highest rated player, Lafferty. Chinook won the USA National Tournament by the widest margin ever. 1995 "No Hands Across America": A semi-autonomous car drove coast-to-coast across the United States with computer-controlled steering for 2,797 miles (4,501km) of the 2,849 miles (4,585km). Throttle and brakes were controlled by a human driver.[41][42] 1995 One of Ernst Dickmanns' robot cars (with robot-controlled throttle and brakes) drove more than 1000 miles from Munich to Copenhagen and back, in traffic, at up to 120mph, occasionally executing maneuvers to pass other cars (only in a few critical situations a safety driver took over). Active vision was used to deal with rapidly changing street scenes. 1997 The Deep Blue chess machine (IBM) defeats the (then) world chess champion, Garry Kasparov. 1997 First official RoboCup football (soccer) match featuring table-top matches with 40 teams of interacting robots and over 5000 spectators. 1997 Computer Othello program Logistello defeated the world champion Takeshi Murakami with a score of 60. 1998 Tiger Electronics' Furby is released, and becomes the first successful attempt at producing a type of A.I to reach a domestic environment. 1998 Tim Berners-Lee published his Semantic Web Road map paper.[43] 1998 Leslie P. Kaelbling, Michael Littman, and Anthony Cassandra introduce the first method for solving POMDPs offline, jumpstarting widespread use in robotics and automated planning and scheduling[44] 1999 Sony introduces an improved domestic robot similar to a Furby, the AIBO becomes one of the first artificially intelligent "pets" that is also autonomous. Late 1990s Web crawlers and other AI-based information extraction programs become essential in widespread use of the World Wide Web. Late 1990s Demonstration of an Intelligent room and Emotional Agents at MIT's AI Lab. Late 1990s Initiation of work on the Oxygen architecture, which connects mobile and stationary computers in an adaptive network. Date Development 2000 Interactive robopets ("smart toys") become commercially available, realizing the vision of the 18th century novelty toy makers. 2000 Cynthia Breazeal at MIT publishes her dissertation on Sociable machines, describing Kismet (robot), with a face that expresses emotions. 2000 The Nomad robot explores remote regions of Antarctica looking for meteorite samples. 2002 iRobot's Roomba autonomously vacuums the floor while navigating and avoiding obstacles. 2004 OWL Web Ontology Language W3C Recommendation (10 February 2004). 2004 DARPA introduces the DARPA Grand Challenge requiring competitors to produce autonomous vehicles for prize money. 2004 NASA's robotic exploration rovers Spirit and Opportunity autonomously navigate the surface of Mars. 2005 Honda's ASIMO robot, an artificially intelligent humanoid robot, is able to walk as fast as a human, delivering trays to customers in restaurant settings. 2005 Recommendation technology based on tracking web activity or media usage brings AI to marketing. See TiVo Suggestions. 2005 Blue Brain is born, a project to simulate the brain at molecular detail.[1] 2006 The Dartmouth Artificial Intelligence Conference: The Next 50 Years (AI@50) AI@50 (1416 July 2006) 2007 Philosophical Transactions of the Royal Society, B Biology, one of the world's oldest scientific journals, puts out a special issue on using AI to understand biological intelligence, titled Models of Natural Action Selection[45] 2007 Checkers is solved by a team of researchers at the University of Alberta. 2007 DARPA launches the Urban Challenge for autonomous cars to obey traffic rules and operate in an urban environment. 2009 Google builds self driving car.[46] Date Development 2010 Microsoft launched Kinect for Xbox 360, the first gaming device to track human body movement, using just a 3D camera and infra-red detection, enabling users to play their Xbox 360 wirelessly. The award-winning machine learning for human motion capture technology for this device was developed by the Computer Vision group at Microsoft Research, Cambridge.[47][48] 2011 IBM's Watson computer defeated television game show Jeopardy! champions Rutter and Jennings. 2011 Apple's Siri, Google's Google Now and Microsoft's Cortana are smartphone apps that use natural language to answer questions, make recommendations and perform actions. 2013 Robot HRP-2 built by SCHAFT Inc of Japan, a subsidiary of Google, defeats 15 teams to win DARPAs Robotics Challenge Trials. HRP-2 scored 27 out of 32 points in 8 tasks needed in disaster response. Tasks are drive a vehicle, walk over debris, climb a ladder, remove debris, walk through doors, cut through a wall, close valves and connect a hose.[49] 2013 NEIL, the Never Ending Image Learner, is released at Carnegie Mellon University to constantly compare and analyze relationships between different images.[50] 2015 An open letter to ban development and use of autonomous weapons signed by Hawking, Musk, Wozniak and 3,000 researchers in AI and robotics.[51] 2015 Google DeepMind's AlphaGo defeated 3 time European Go champion 2 dan professional Fan Hui by 5 games to 0.[52] 2016 Google DeepMind's AlphaGo defeated Lee Sedol 4-1. Lee Sedol is a 9 dan professional Korean Go champion who won 27 major tournaments from 2002 to 2016.[53] Before the match with AlphaGo, Lee Sedol was confident in predicting an easy 5-0 or 4-1 victory.[54] 2017 Google DeepMind's AlphaGo won 60-0 rounds on two public Go websites including 3 wins against world Go champion Ke Jie. [55] 2017 Libratus, designed by Carnegie Mellon professor Tuomas Sandholm and his grad student Noam Brown won against four top players at no-limit Texas hold 'em, a very challenging version of poker. Unlike Go and Chess, Poker is a game in which some information is hidden (the cards of the other player) which makes it much harder to model. [56]

Read the rest here:

Timeline of artificial intelligence - Wikipedia

Artificial Intelligence ('AI') is a runaway success and we think it is going to be one of the biggest, if not the biggest driver of future economic growth. There are major AI breakthroughs on a fundamental level leading to a host of groundbreaking applications in autonomous driving, medical diagnostics, automatic translation, speech recognition and a host more.

See for instance the acceleration in speech recognition in the last year or so:

We're only at the beginning of these developments, which is going in several overlapping stages:

We have described the development of specialist AI chips in an earlier article, where we already touched on the new phase emerging - the move of AI from the cloud to the device (usually the mobile phone).

This certainly isn't a universal movement but involves inference (the application of the algorithms to answer queries), rather than the more computing-heavy training (where the algorithms are improved through iteration rounds with the help of massive amounts of data).

Since GPUs weren't designed with AI in mind, so in principle, it isn't much of a stretch to assume that specialist AI chips will take performance higher, even if Nvidia is now designing new architectures like the Volta with AI in mind at least in part, from Medium:

Although Pascal has performed well in deep learning, Volta is far superior because it unifies CUDA Cores and Tensor Cores. Tensor Cores are a breakthrough technology designed to speed up AI workloads. The Volta Tensor Cores can generate 12 times more throughput than Pascal, allowing the Tesla V100 to deliver 120 teraflops (a measure of GPU power) of deep learning performance... The new Volta-powered DGX-1 leapfrogs its previous version with significant advances in TFLOPS (170 to 960), CUDA cores (28,672 to 40,960), Tensor Cores (0 to 5120), NVLink vs. PCIe speed-up (5X to 10X), and deep learning training speed (1X to 3X).

However, while the systems on a chip (SoC) that drive mobile devices contain a GPU processor, these are pretty tiny compared to their desktop and server equivalents. There is room here too for adding intelligence locally (or, as the jargon has it, 'on the edge').

Advantages

Why would one want to put AI processing 'on the edge' (on the device rather than in the cloud)? There are a few reasons:

The privacy issue was best explained by SA contributor Mark Hibben:

The motivation for this is customer privacy. Currently, AI assistants such as Siri, Cortana, Google Assistant, and Alexa are all hosted in the cloud and require Internet connections to access. The simple reason for this is that AI functionality requires a lot of processing horsepower that only datacenters could provide. But this constitutes a potential privacy issue for users, since cloud-hosted AIs are most effective when they are observing the actions of the user. That way they can learn the users' needs and be more "assistive". This means that virtually every user action, including voice and text messaging, could be subject to such observation. This has prompted Apple to look for ways to host some AI functionality on the mobile device, where it can be locked behind the protection of Apple's redoubtable Secure Enclave. The barrier to this is simply the magnitude of the processing task.

Lower latency and a possible lack of internet connection are crucial where there are life and death decisions that have to be taken instantly, for instance in autonomous driving.

Security of devices might benefit from AI-driven behavioural malware applications, which could run more efficient on specialist chips locally, rather than via the cloud.

Specialist AI chips might also provide an energy advantage, especially when some AI applications already use the local resources (CPU, GPU), and/or depend for data on the cloud (especially in scenarios where there is no Wi-Fi available). We understand that this is one motivation for Apple (NASDAQ:AAPL) to develop its own AI chips.

But here are some of the challenges, very well explained by Google (NASDAQ:GOOG) (NASDAQ:GOOGL):

These low-end phones can be about 50 times slower than a good laptop-and a good laptop is already much slower than the data centers that typically run our image recognition systems. So how do we get visual translation on these phones, with no connection to the cloud, translating in real-time as the camera moves around? We needed to develop a very small neural net, and put severe limits on how much we tried to teach it-in essence, put an upper bound on the density of information it handles. The challenge here was in creating the most effective training data. Since we're generating our own training data, we put a lot of effort into including just the right data and nothing more.

One route is what Google is doing by optimizing these very small neural nets and feeding it with just the right amount of data. However, if more resources were available locally on the device, these constraints would loosen. Hence, the search for a mobile AI chip that is more efficient in handling these neural networks.

ARM

ARM, now part of the Japanese SoftBank (OTCPK:SFTBY), is adapting its architecture to produce better results for AI. For instance, its DynamiQ architecture, from The Verge:

Dynamiq goes beyond offering just additional flexibility, and will also let chip makers optimize their silicon for tasks like machine learning. Companies will have the option of building AI accelerators directly into chips, helping systems manage data and memory more efficiently. These accelerators could mean that machine learning-powered software features (like Huawei's latest OS, which studies the apps users use most and allocates processing power accordingly) could be implemented more efficiently.

ARM is claiming that DynamiQ will deliver a 50 times increase in "AI-related performance" over the next three to five years. That remains to be seen, but it's noteworthy that they are designing chips with AI in mind.

Qualcomm (NASDAQ:QCOM)

The major user of ARM designs is Qualcomm and this company is also adding AI capabilities to its chips. It isn't adding hardware, but a machine learning platform called Zeroth, or the Snapdragon Neural Processing Engine.

It's a software development kit that will make it easier to develop deep learning programs directly on the mobile (and other devices run by Snapdragon processors). Here is the selling point ( The Verge):

This means that if companies want to build their own deep learning analytics, they won't have to rent servers to deliver their software to customers. And although running deep learning operations locally means limiting their complexity, the sort of programs you can run on your phone or any other portable device are still impressive. The real limitation will be Qualcomm's chips. The new SDK will only work with the latest Snapdragon 820 processors from the latter half of 2016, and the company isn't saying if it plans to expand its availability.

Snapdragons like the 825, the flagship 835 and some of the 600-tier chips incorporate some machine learning capabilities. And they're not doing this all by themselves either, from Qualcomm:

An exciting development in this field is Facebook's stepped up investment in Caffe2, the evolution of the open source Caffe framework. At this year's F8 conference, Facebook and Qualcomm Technologies announced a collaboration to support the optimization of Caffe2, Facebook's open source deep learning framework, and the Qualcomm Snapdragon neural processing engine (NPE) framework. The NPE is designed to do the heavy lifting needed to run neural networks efficiently on Snapdragon, leaving developers with more time and resources to focus on creating their innovative user experiences.

IBM (NYSE:IBM)

IBM is developing its own specialist AI chip called True North. It is a unique product that mirrors the design of neural networks. It will be like a 'brain on a phone' the size of the brain of a small rodent, packing 48 million electronic nerve cells, from Wired:

Each chip mimics about a million neurons, and these can communicate with each other via something similar to a synapse, the connections between neurons in the brain.

The chip won't be out for quite some time, but its main benefit is that it's exceptionally frugal, from Wired:

The upshot is a much simpler architecture that consumes less power. Though the chip contains 5.4 billion transistors, it draws about 70 milliwatts of power. A standard Intel computer processor, by comparison, includes 1.4 billion transistors and consumes about 35 to 140 watts. Even the ARM chips that drive smartphones consume several times more power than the TrueNorth.

For now, it will do the less computationally heavy stuff involved in inferencing, not the training part of machine learning (feeding algorithms massive amounts of data in order to improve them). From Wired:

But the promise is that IBM's chip can run these algorithms in smaller spaces with considerably less electrical power, letting us shoehorn more AI onto phones and other tiny devices, including hearing aids and, well, wristwatches.

Considering its energy needs, IBM's True North is perhaps the prime candidate to add local intelligence to devices, even tiny ones. This could ultimately revolutionize the internet of things (IoT), which itself is still in its infancy but based on simple processors and sensors.

Adding intelligence to IoT devices and interconnecting these opens up distributed computing on a staggering scale, but speculation about its possibilities is best left for another time.

Apple

Apple is also working on an AI chip for mobile devices, Apple's Neural Engine. There isn't much known in terms of detail; its use is to offload tasks from the CPU and GPU so saving battery and speed up stuff like face and speech recognition and mixed reality.

Groq

Then there is the startup called Groq, founded by some of the people who developed the Tensor at Google. Unfortunately, at this stage, there is very little known about the company, apart from the fact that they're developing a Tensor like AI chip. Here is Venture capitalist Chamath Palihapitiya (from CNBC):

There are no promotional materials or website. All that exists online are a couple SEC filings from October and December showing that the company raised $10.3 million, and an incorporation filing in the state of Delaware on Sept. 12. "We're really excited about Groq," Palihapitiya wrote in an e-mail. "It's too early to talk specifics, but we think what they're building could become a fundamental building block for the next generation of computing."

It's certainly a daring venture as the cost of erecting a new chip company from scratch can be exorbitant and the company faces well established competitors with Google, Apple and Nvidia (NASDAQ:NVDA).

What is also unknown is whether the chip is for datacenters or smaller devices providing local AI processing.

Nvidia

The current leader for datacenter "AI" chips (obviously, these are not specific AI chips but GPUs that are used to do most of the massive parallel computing of training neural networks to improve the accuracy of the algorithms.

But it is building its own solution for local AI computing in the form of the Xavier SoC, integrating CPU, CUDA GPU and deep learning accelerators and the GPU now contains the new Volta architecture. It is built for the forthcoming Drive PX3 (autonomous driving).

However, Nvidia's Xavier will feature its own form of TPU that it calls a Tensor Core, and this is built into the SOC.

The advantage for on-device computing in autonomous driving is clear - it reduces latency and the risk of loss of internet connection. Critical autonomous driving functions simply cannot rely on spotty internet connections or long latencies.

From what we understand, it's like a supercomputer in a box, but that's still too big (and too power hungry, sipping 20W) for smartphones. But needless to say, autonomous driving is a big emerging market in and by itself, and in time, this stuff tends to miniaturize, and that TPU itself will be a lot smaller and less energy hungry so it might very well be applicable in other environments.

Conclusion

Before we get too excited, there are serious limitations to putting too much AI computing on small devices like smartphones, here is Voicebot:

The third chip approach seems logical for on-device AI processing. However, few AI processes actually occur on-device today. Whether it is Amazon's Alexa or Apple's Siri, the language processing and understanding occurs in the cloud. It would be impressive if Apple could actually bring all of Siri's language understanding processing onto a mobile device, but that is unlikely in the near term. It's not just about analyzing the data, it's also about having access to information that helps you interpret and respond to requests. The cloud is well suited to these challenges.

Most AI requires massive amounts of computing power and massive amounts of data. While some of that can be shifted from the cloud to devices, especially where latency and secure coverage are essential (autonomous driving), there are still significant limitations for what can be done locally.

However, the development of specialist AI chips for local (rather than cloud) use is only starting today and a new and exciting market is opening up here, with big companies like Apple, Nvidia, STMicroelectronics (NYSE:STM), and IBM all at it. And the companies developing cloud AI chips, like Google and Groq might very well crack this market too, as Google's Tensor seems particularly efficient in terms of energy use.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Original post:

Artificial Intelligence: From The Cloud To Your Pocket - Seeking Alpha