Daily Archives: March 23, 2017

Artificial intelligence (AI) is intelligence exhibited by machines. In computer science, the field of AI research defines itself as the study of "intelligent agents": any device that perceives its environment and takes actions that maximize its chance of success at some goal.[1] Colloquially, the term "artificial intelligence" is applied when a machine mimics "cognitive" functions that humans associate with other human minds, such as "learning" and "problem solving" (known as Machine Learning). As machines become increasingly capable, mental facilities once thought to require intelligence are removed from the definition. For instance, optical character recognition is no longer perceived as an exemplar of "artificial intelligence", having become a routine technology.[3] Capabilities currently classified as AI include successfully understanding human speech, competing at a high level in strategic game systems (such as Chess and Go[5]), self-driving cars, intelligent routing in content delivery networks, and interpreting complex data.

AI research is divided into subfields[6] that focus on specific problems or on specific approaches or on the use of a particular tool or towards satisfying particular applications.

The central problems (or goals) of AI research include reasoning, knowledge, planning, learning, natural language processing (communication), perception and the ability to move and manipulate objects.[7]General intelligence is among the field's long-term goals.[8] Approaches include statistical methods, computational intelligence, and traditional symbolic AI. Many tools are used in AI, including versions of search and mathematical optimization, logic, methods based on probability and economics. The AI field draws upon computer science, mathematics, psychology, linguistics, philosophy, neuroscience and artificial psychology.

The field was founded on the claim that human intelligence "can be so precisely described that a machine can be made to simulate it".[9] This raises philosophical arguments about the nature of the mind and the ethics of creating artificial beings endowed with human-like intelligence, issues which have been explored by myth, fiction and philosophy since antiquity.[10] Some people also consider AI a danger to humanity if it progresses unabatedly.[11] Attempts to create artificial intelligence have experienced many setbacks, including the ALPAC report of 1966, the abandonment of perceptrons in 1970, the Lighthill Report of 1973, the second AI winter 19871993 and the collapse of the Lisp machine market in 1987.

In the twenty-first century, AI techniques, both "hard" and "soft" have experienced a resurgence following concurrent advances in computer power, sizes of training sets, and theoretical understanding, and AI techniques have become an essential part of the technology industry, helping to solve many challenging problems in computer science.[12]

While thought-capable artificial beings appeared as storytelling devices in antiquity,[13] the idea of actually trying to build a machine to perform useful reasoning may have begun with Ramon Llull (c. 1300 CE). With his Calculus ratiocinator, Gottfried Leibniz extended the concept of the calculating machine (Wilhelm Schickard engineered the first one around 1623), intending to perform operations on concepts rather than numbers. Since the 19th century, artificial beings are common in fiction, as in Mary Shelley's Frankenstein or Karel apek's R.U.R. (Rossum's Universal Robots).[15]

The study of mechanical or "formal" reasoning began with philosophers and mathematicians in antiquity. In the 19th century, George Boole refined those ideas into propositional logic and Gottlob Frege developed a notational system for mechanical reasoning (a "predicate calculus"). Around the 1940s, Alan Turing's theory of computation suggested that a machine, by shuffling symbols as simple as "0" and "1", could simulate any conceivable act of mathematical deduction. This insight, that digital computers can simulate any process of formal reasoning, is known as the ChurchTuring thesis.[17][pageneeded] Along with concurrent discoveries in neurology, information theory and cybernetics, this led researchers to consider the possibility of building an electronic brain.[18] The first work that is now generally recognized as AI was McCullouch and Pitts' 1943 formal design for Turing-complete "artificial neurons".

The field of AI research was "born"[19] at a conference at Dartmouth College in 1956.[20] Attendees Allen Newell (CMU), Herbert Simon (CMU), John McCarthy (MIT), Marvin Minsky (MIT) and Arthur Samuel (IBM) became the founders and leaders of AI research.[21] At the conference, Newell and Simon, together with programmer J. C. Shaw (RAND), presented the first true artificial intelligence program, the Logic Theorist. This spurred tremendous research in the domain: computers were winning at checkers, solving word problems in algebra, proving logical theorems and speaking English.[23] By the middle of the 1960s, research in the U.S. was heavily funded by the Department of Defense[24] and laboratories had been established around the world.[25] AI's founders were optimistic about the future: Herbert Simon predicted, "machines will be capable, within twenty years, of doing any work a man can do." Marvin Minsky agreed, writing, "within a generation... the problem of creating 'artificial intelligence' will substantially be solved."[26]

They failed to recognize the difficulty of some of the remaining tasks. Progress slowed and in 1974, in response to the criticism of Sir James Lighthill and ongoing pressure from the US Congress to fund more productive projects, both the U.S. and British governments cut off exploratory research in AI. The next few years would later be called an "AI winter",[28] a period when funding for AI projects was hard to find.

In the early 1980s, AI research was revived by the commercial success of expert systems,[29] a form of AI program that simulated the knowledge and analytical skills of human experts. By 1985 the market for AI had reached over a billion dollars. At the same time, Japan's fifth generation computer project inspired the U.S and British governments to restore funding for academic research.[30] However, beginning with the collapse of the Lisp Machine market in 1987, AI once again fell into disrepute, and a second, longer-lasting hiatus began.[31]

In the late 1990s and early 21st century, AI began to be used for logistics, data mining, medical diagnosis and other areas.[12] The success was due to increasing computational power (see Moore's law), greater emphasis on solving specific problems, new ties between AI and other fields and a commitment by researchers to mathematical methods and scientific standards.[32]Deep Blue became the first computer chess-playing system to beat a reigning world chess champion, Garry Kasparov on 11 May 1997.

Advanced statistical techniques (loosely known as deep learning), access to large amounts of data and faster computers enabled advances in machine learning and perception.[34] By the mid 2010s, machine learning applications were used throughout the world.[35] In a Jeopardy! quiz show exhibition match, IBM's question answering system, Watson, defeated the two greatest Jeopardy champions, Brad Rutter and Ken Jennings, by a significant margin. The Kinect, which provides a 3D bodymotion interface for the Xbox 360 and the Xbox One use algorithms that emerged from lengthy AI research[37] as do intelligent personal assistants in smartphones.[38] In March 2016, AlphaGo won 4 out of 5 games of Go in a match with Go champion Lee Sedol, becoming the first computer Go-playing system to beat a professional Go player without handicaps.[5][39]

According to Bloomberg's Jack Clark, 2015 was a landmark year for artificial intelligence, with the number of software projects that use AI within Google increasing from a "sporadic usage" in 2012 to more than 2,700 projects. Clark also presents factual data indicating that error rates in image processing tasks have fallen significantly since 2011.[40] He attributes this to an increase in affordable neural networks, due to a rise in cloud computing infrastructure and to an increase in research tools and datasets. Other cited examples include Microsoft's development of a Skype system that can automatically translate from one language to another and Facebook's system that can describe images to blind people.[40]

The overall research goal of artificial intelligence is to create technology that allows computers and machines to function in an intelligent manner. The general problem of simulating (or creating) intelligence has been broken down into sub-problems. These consist of particular traits or capabilities that researchers expect an intelligent system to display. The traits described below have received the most attention.[7]

Erik Sandwell emphasizes planning and learning that is relevant and applicable to the given situation.[41]

Early researchers developed algorithms that imitated step-by-step reasoning that humans use when they solve puzzles or make logical deductions (reason).[42] By the late 1980s and 1990s, AI research had developed methods for dealing with uncertain or incomplete information, employing concepts from probability and economics.[43]

For difficult problems, algorithms can require enormous computational resourcesmost experience a "combinatorial explosion": the amount of memory or computer time required becomes astronomical for problems of a certain size. The search for more efficient problem-solving algorithms is a high priority.[44]

Human beings ordinarily use fast, intuitive judgments rather than step-by-step deduction that early AI research was able to model.[45] AI has progressed using "sub-symbolic" problem solving: embodied agent approaches emphasize the importance of sensorimotor skills to higher reasoning; neural net research attempts to simulate the structures inside the brain that give rise to this skill; statistical approaches to AI mimic the human ability.

Knowledge representation[46] and knowledge engineering[47] are central to AI research. Many of the problems machines are expected to solve will require extensive knowledge about the world. Among the things that AI needs to represent are: objects, properties, categories and relations between objects;[48] situations, events, states and time;[49] causes and effects;[50] knowledge about knowledge (what we know about what other people know);[51] and many other, less well researched domains. A representation of "what exists" is an ontology: the set of objects, relations, concepts and so on that the machine knows about. The most general are called upper ontologies, which attempt to provide a foundation for all other knowledge.[52]

Among the most difficult problems in knowledge representation are:

Intelligent agents must be able to set goals and achieve them.[59] They need a way to visualize the future (they must have a representation of the state of the world and be able to make predictions about how their actions will change it) and be able to make choices that maximize the utility (or "value") of the available choices.[60]

In classical planning problems, the agent can assume that it is the only thing acting on the world and it can be certain what the consequences of its actions may be.[61] However, if the agent is not the only actor, it must periodically ascertain whether the world matches its predictions and it must change its plan as this becomes necessary, requiring the agent to reason under uncertainty.[62]

Multi-agent planning uses the cooperation and competition of many agents to achieve a given goal. Emergent behavior such as this is used by evolutionary algorithms and swarm intelligence.[63]

Machine learning is the study of computer algorithms that improve automatically through experience[64][65] and has been central to AI research since the field's inception.[66]

Unsupervised learning is the ability to find patterns in a stream of input. Supervised learning includes both classification and numerical regression. Classification is used to determine what category something belongs in, after seeing a number of examples of things from several categories. Regression is the attempt to produce a function that describes the relationship between inputs and outputs and predicts how the outputs should change as the inputs change. In reinforcement learning[67] the agent is rewarded for good responses and punished for bad ones. The agent uses this sequence of rewards and punishments to form a strategy for operating in its problem space. These three types of learning can be analyzed in terms of decision theory, using concepts like utility. The mathematical analysis of machine learning algorithms and their performance is a branch of theoretical computer science known as computational learning theory.[68]

Within developmental robotics, developmental learning approaches were elaborated for lifelong cumulative acquisition of repertoires of novel skills by a robot, through autonomous self-exploration and social interaction with human teachers, and using guidance mechanisms such as active learning, maturation, motor synergies, and imitation.[69][70]

Natural language processing[73] gives machines the ability to read and understand the languages that humans speak. A sufficiently powerful natural language processing system would enable natural language user interfaces and the acquisition of knowledge directly from human-written sources, such as newswire texts. Some straightforward applications of natural language processing include information retrieval, text mining, question answering[74] and machine translation.[75]

A common method of processing and extracting meaning from natural language is through semantic indexing. Increases in processing speeds and the drop in the cost of data storage makes indexing large volumes of abstractions of the user's input much more efficient.

Machine perception[76] is the ability to use input from sensors (such as cameras, microphones, tactile sensors, sonar and others more exotic) to deduce aspects of the world. Computer vision[77] is the ability to analyze visual input. A few selected subproblems are speech recognition,[78]facial recognition and object recognition.[79]

The field of robotics[80] is closely related to AI. Intelligence is required for robots to be able to handle such tasks as object manipulation[81] and navigation, with sub-problems of localization (knowing where you are, or finding out where other things are), mapping (learning what is around you, building a map of the environment), and motion planning (figuring out how to get there) or path planning (going from one point in space to another point, which may involve compliant motion where the robot moves while maintaining physical contact with an object).[83]

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 sciences, psychology, and cognitive science. While the origins of the field may be traced as far back as to early philosophical inquiries into emotion, the more modern branch of computer science originated with Rosalind Picard's 1995 paper on affective computing.[90][91] A motivation for the research is the ability to simulate empathy. The machine should interpret the emotional state of humans and adapt its behaviour to them, giving an appropriate response for those emotions.

Emotion and social skills[92] play two roles for an intelligent agent. First, it must be able to predict the actions of others, by understanding their motives and emotional states. (This involves elements of game theory, decision theory, as well as the ability to model human emotions and the perceptual skills to detect emotions.) Also, in an effort to facilitate human-computer interaction, an intelligent machine might want to be able to display emotionseven if it does not actually experience them itselfin order to appear sensitive to the emotional dynamics of human interaction.

A sub-field of AI addresses creativity both theoretically (from a philosophical and psychological perspective) and practically (via specific implementations of systems that generate outputs that can be considered creative, or systems that identify and assess creativity). Related areas of computational research are Artificial intuition and Artificial thinking.

Many researchers think that their work will eventually be incorporated into a machine with artificial general intelligence, combining all the skills above and exceeding human abilities at most or all of them.[8][93] A few believe that anthropomorphic features like artificial consciousness or an artificial brain may be required for such a project.[94][95]

Many of the problems above may require general intelligence to be considered solved. For example, even a straightforward, specific task like machine translation requires that the machine read and write in both languages (NLP), follow the author's argument (reason), know what is being talked about (knowledge), and faithfully reproduce the author's intention (social intelligence). A problem like machine translation is considered "AI-complete". In order to reach human-level performance for machines, one must solve all the problems.[96]

There is no established unifying theory or paradigm that guides AI research. Researchers disagree about many issues.[97] A few of the most long standing questions that have remained unanswered are these: should artificial intelligence simulate natural intelligence by studying psychology or neurology? Or is human biology as irrelevant to AI research as bird biology is to aeronautical engineering?[98] Can intelligent behavior be described using simple, elegant principles (such as logic or optimization)? Or does it necessarily require solving a large number of completely unrelated problems?[99] Can intelligence be reproduced using high-level symbols, similar to words and ideas? Or does it require "sub-symbolic" processing?[100] John Haugeland, who coined the term GOFAI (Good Old-Fashioned Artificial Intelligence), also proposed that AI should more properly be referred to as synthetic intelligence,[101] a term which has since been adopted by some non-GOFAI researchers.

Stuart Shapiro divides AI research into three approaches, which he calls computational psychology, computational philosophy, and computer science. Computational psychology is used to make computer programs that mimic human behavior.[104] Computational philosophy, is used to develop an adaptive, free-flowing computer mind.[104] Implementing computer science serves the goal of creating computers that can perform tasks that only people could previously accomplish.[104] Together, the humanesque behavior, mind, and actions make up artificial intelligence.

In the 1940s and 1950s, a number of researchers explored the connection between neurology, information theory, and cybernetics. Some of them built machines that used electronic networks to exhibit rudimentary intelligence, such as W. Grey Walter's turtles and the Johns Hopkins Beast. Many of these researchers gathered for meetings of the Teleological Society at Princeton University and the Ratio Club in England.[18] By 1960, this approach was largely abandoned, although elements of it would be revived in the 1980s.

When access to digital computers became possible in the middle 1950s, AI research began to explore the possibility that human intelligence could be reduced to symbol manipulation. The research was centered in three institutions: Carnegie Mellon University, Stanford and MIT, and each one developed its own style of research. John Haugeland named these approaches to AI "good old fashioned AI" or "GOFAI".[105] During the 1960s, symbolic approaches had achieved great success at simulating high-level thinking in small demonstration programs. Approaches based on cybernetics or neural networks were abandoned or pushed into the background.[106] Researchers in the 1960s and the 1970s were convinced that symbolic approaches would eventually succeed in creating a machine with artificial general intelligence and considered this the goal of their field.

By the 1980s progress in symbolic AI seemed to stall and many believed that symbolic systems would never be able to imitate all the processes of human cognition, especially perception, robotics, learning and pattern recognition. A number of researchers began to look into "sub-symbolic" approaches to specific AI problems.[100] Sub-symbolic methods manage to approach intelligence without specific representations of knowledge.

In the 1990s, AI researchers developed sophisticated mathematical tools to solve specific subproblems. These tools are truly scientific, in the sense that their results are both measurable and verifiable, and they have been responsible for many of AI's recent successes. The shared mathematical language has also permitted a high level of collaboration with more established fields (like mathematics, economics or operations research). Stuart Russell and Peter Norvig describe this movement as nothing less than a "revolution" and "the victory of the neats".[32] Critics argue that these techniques (with few exceptions) are too focused on particular problems and have failed to address the long-term goal of general intelligence. There is an ongoing debate about the relevance and validity of statistical approaches in AI, exemplified in part by exchanges between Peter Norvig and Noam Chomsky.

In the course of 50 years of research, AI has developed a large number of tools to solve the most difficult problems in computer science. A few of the most general of these methods are discussed below.

Many problems in AI can be solved in theory by intelligently searching through many possible solutions:[124]Reasoning can be reduced to performing a search. For example, logical proof can be viewed as searching for a path that leads from premises to conclusions, where each step is the application of an inference rule.[125]Planning algorithms search through trees of goals and subgoals, attempting to find a path to a target goal, a process called means-ends analysis.[126]Robotics algorithms for moving limbs and grasping objects use local searches in configuration space.[81] Many learning algorithms use search algorithms based on optimization.

Simple exhaustive searches[127] 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 or never completes. The solution, for many problems, is to use "heuristics" or "rules of thumb" that eliminate choices that are unlikely to lead to the goal (called "pruning the search tree"). Heuristics supply the program with a "best guess" for the path on which the solution lies.[128] Heuristics limit the search for solutions into a smaller sample size.

A very different kind of search came to prominence in the 1990s, based on the mathematical theory of optimization. For many problems, it is possible to begin the search with some form of a guess and then refine the guess incrementally until no more refinements can be made. These algorithms can be visualized as blind hill climbing: we begin the search at a random point on the landscape, and then, by jumps or steps, we keep moving our guess uphill, until we reach the top. Other optimization algorithms are simulated annealing, beam search and random optimization.[129]

Evolutionary computation uses a form of optimization search. For example, they may begin with a population of organisms (the guesses) and then allow them to mutate and recombine, selecting only the fittest to survive each generation (refining the guesses). Forms of evolutionary computation include swarm intelligence algorithms (such as ant colony or particle swarm optimization)[130] and evolutionary algorithms (such as genetic algorithms, gene expression programming, and genetic programming).[131]

Logic[132] is used for knowledge representation and problem solving, but it can be applied to other problems as well. For example, the satplan algorithm uses logic for planning[133] and inductive logic programming is a method for learning.[134]

Several different forms of logic are used in AI research. Propositional or sentential logic[135] is the logic of statements which can be true or false. First-order logic[136] also allows the use of quantifiers and predicates, and can express facts about objects, their properties, and their relations with each other. Fuzzy logic,[137] is a version of first-order logic which allows the truth of a statement to be represented as a value between 0 and 1, rather than simply True (1) or False (0). Fuzzy systems can be used for uncertain reasoning and have been widely used in modern industrial and consumer product control systems. Subjective logic[138] models uncertainty in a different and more explicit manner than fuzzy-logic: a given binomial opinion satisfies belief + disbelief + uncertainty = 1 within a Beta distribution. By this method, ignorance can be distinguished from probabilistic statements that an agent makes with high confidence.

Default logics, non-monotonic logics and circumscription[54] are forms of logic designed to help with default reasoning and the qualification problem. Several extensions of logic have been designed to handle specific domains of knowledge, such as: description logics;[48]situation calculus, event calculus and fluent calculus (for representing events and time);[49]causal calculus;[50] belief calculus;[139] and modal logics.[51]

Many problems in AI (in reasoning, planning, learning, perception and robotics) require the agent to operate with incomplete or uncertain information. AI researchers have devised a number of powerful tools to solve these problems using methods from probability theory and economics.[140]

Bayesian networks[141] are a very general tool that can be used for a large number of problems: reasoning (using the Bayesian inference algorithm),[142]learning (using the expectation-maximization algorithm),[143]planning (using decision networks)[144] and perception (using dynamic Bayesian networks).[145] Probabilistic algorithms can also be used for filtering, prediction, smoothing and finding explanations for streams of data, helping perception systems to analyze processes that occur over time (e.g., hidden Markov models or Kalman filters).[145]

A key concept from the science of economics is "utility": a measure of how valuable something is to an intelligent agent. Precise mathematical tools have been developed that analyze how an agent can make choices and plan, using decision theory, decision analysis,[146] and information value theory.[60] These tools include models such as Markov decision processes,[147] dynamic decision networks,[145]game theory and mechanism design.[148]

The simplest AI applications can be divided into two types: classifiers ("if shiny then diamond") and controllers ("if shiny then pick up"). Controllers do, however, also classify conditions before inferring actions, and therefore classification forms a central part of many AI systems. Classifiers are functions that use pattern matching to determine a closest match. They can be tuned according to examples, making them very attractive for use in AI. These examples are known as observations or patterns. In supervised learning, each pattern belongs to a certain predefined class. A class can be seen as a decision that has to be made. All the observations combined with their class labels are known as a data set. When a new observation is received, that observation is classified based on previous experience.[149]

A classifier can be trained in various ways; there are many statistical and machine learning approaches. The most widely used classifiers are the neural network,[150]kernel methods such as the support vector machine,[151]k-nearest neighbor algorithm,[152]Gaussian mixture model,[153]naive Bayes classifier,[154] and decision tree.[155] The performance of these classifiers have been compared over a wide range of tasks. Classifier performance depends greatly on the characteristics of the data to be classified. There is no single classifier that works best on all given problems; this is also referred to as the "no free lunch" theorem. Determining a suitable classifier for a given problem is still more an art than science.[156]

The study of non-learning artificial neural networks[150] began in the decade before the field of AI research was founded, in the work of Walter Pitts and Warren McCullouch. Frank Rosenblatt invented the perceptron, a learning network with a single layer, similar to the old concept of linear regression. Early pioneers also include Alexey Grigorevich Ivakhnenko, Teuvo Kohonen, Stephen Grossberg, Kunihiko Fukushima, Christoph von der Malsburg, David Willshaw, Shun-Ichi Amari, Bernard Widrow, John Hopfield, Eduardo R. Caianiello, and others.

The main categories of networks are acyclic or feedforward neural networks (where the signal passes in only one direction) and recurrent neural networks (which allow feedback and short-term memories of previous input events). Among the most popular feedforward networks are perceptrons, multi-layer perceptrons and radial basis networks.[157] Neural networks can be applied to the problem of intelligent control (for robotics) or learning, using such techniques as Hebbian learning, GMDH or competitive learning.[158]

Today, neural networks are often trained by the backpropagation algorithm, which had been around since 1970 as the reverse mode of automatic differentiation published by Seppo Linnainmaa,[159][160] and was introduced to neural networks by Paul Werbos.[161][162][163]

Hierarchical temporal memory is an approach that models some of the structural and algorithmic properties of the neocortex.[164]

Deep learning in artificial neural networks with many layers has transformed many important subfields of artificial intelligence, including computer vision, speech recognition, natural language processing and others.[165][166][167]

According to a survey,[168] the expression "Deep Learning" was introduced to the Machine Learning community by Rina Dechter in 1986[169] and gained traction after Igor Aizenberg and colleagues introduced it to Artificial Neural Networks in 2000.[170] The first functional Deep Learning networks were published by Alexey Grigorevich Ivakhnenko and V. G. Lapa in 1965.[171][pageneeded] These networks are trained one layer at a time. Ivakhnenko's 1971 paper[172] describes the learning of a deep feedforward multilayer perceptron with eight layers, already much deeper than many later networks. In 2006, a publication by Geoffrey Hinton and Ruslan Salakhutdinov introduced another way of pre-training many-layered feedforward neural networks (FNNs) one layer at a time, treating each layer in turn as an unsupervised restricted Boltzmann machine, then using supervised backpropagation for fine-tuning. Similar to shallow artificial neural networks, deep neural networks can model complex non-linear relationships. Over the last few years, advances in both machine learning algorithms and computer hardware have led to more efficient methods for training deep neural networks that contain many layers of non-linear hidden units and a very large output layer.[174]

Deep learning often uses convolutional neural networks (CNNs), whose origins can be traced back to the Neocognitron introduced by Kunihiko Fukushima in 1980.[175] In 1989, Yann LeCun and colleagues applied backpropagation to such an architecture. In the early 2000s, in an industrial application CNNs already processed an estimated 10% to 20% of all the checks written in the US.[176] Since 2011, fast implementations of CNNs on GPUs have won many visual pattern recognition competitions.[167]

Deep feedforward neural networks were used in conjunction with reinforcement learning by AlphaGo, Google Deepmind's program that was the first to beat a professional human player.[177]

Early on, deep learning was also applied to sequence learning with recurrent neural networks (RNNs)[178] which are general computers and can run arbitrary programs to process arbitrary sequences of inputs. The depth of an RNN is unlimited and depends on the length of its input sequence.[167] RNNs can be trained by gradient descent[179][180][181] but suffer from the vanishing gradient problem.[165][182] In 1992, it was shown that unsupervised pre-training of a stack of recurrent neural networks can speed up subsequent supervised learning of deep sequential problems.[183]

Numerous researchers now use variants of a deep learning recurrent NN called the long short-term memory (LSTM) network published by Hochreiter & Schmidhuber in 1997.[184] LSTM is often trained by Connectionist Temporal Classification (CTC).[185] At Google, Microsoft and Baidu this approach has revolutionised speech recognition.[186][187][188] For example, in 2015, Google's speech recognition experienced a dramatic performance jump of 49% through CTC-trained LSTM, which is now available through Google Voice to billions of smartphone users.[189] Google also used LSTM to improve machine translation,[190] Language Modeling[191] and Multilingual Language Processing.[192] LSTM combined with CNNs also improved automatic image captioning[193] and a plethora of other applications.

Control theory, the grandchild of cybernetics, has many important applications, especially in robotics.[194]

AI researchers have developed several specialized languages for AI research, including Lisp[195] and Prolog.[196]

In 1950, Alan Turing proposed a general procedure to test the intelligence of an agent now known as the Turing test. This procedure allows almost all the major problems of artificial intelligence to be tested. However, it is a very difficult challenge and at present all agents fail.[197]

Artificial intelligence can also be evaluated on specific problems such as small problems in chemistry, hand-writing recognition and game-playing. Such tests have been termed subject matter expert Turing tests. Smaller problems provide more achievable goals and there are an ever-increasing number of positive results.[198]

For example, performance at draughts (i.e. checkers) is optimal,[199] performance at chess is high-human and nearing super-human (see computer chess:computers versus human) and performance at many everyday tasks (such as recognizing a face or crossing a room without bumping into something) is sub-human.

A quite different approach measures machine intelligence through tests which are developed from mathematical definitions of intelligence. Examples of these kinds of tests start in the late nineties devising intelligence tests using notions from Kolmogorov complexity and data compression.[200] Two major advantages of mathematical definitions are their applicability to nonhuman intelligences and their absence of a requirement for human testers.

A derivative of the Turing test is the Completely Automated Public Turing test to tell Computers and Humans Apart (CAPTCHA). As the name implies, this helps to determine that a user is an actual person and not a computer posing as a human. In contrast to the standard Turing test, CAPTCHA administered by a machine and targeted to a human as opposed to being administered by a human and targeted to a machine. A computer asks a user to complete a simple test then generates a grade for that test. Computers are unable to solve the problem, so correct solutions are deemed to be the result of a person taking the test. A common type of CAPTCHA is the test that requires the typing of distorted letters, numbers or symbols that appear in an image undecipherable by a computer.

AI is relevant to any intellectual task. Modern artificial intelligence techniques are pervasive and are too numerous to list here. Frequently, when a technique reaches mainstream use, it is no longer considered artificial intelligence; this phenomenon is described as the AI effect.

High-profile examples of AI include autonomous vehicles (such as drones and self-driving cars), medical diagnosis, creating art (such as poetry), proving mathematical theorems, playing games (such as Chess or Go), search engines (such as Google search), online assistants (such as Siri), image recognition in photographs, spam filtering, prediction of judicial decisions[204] and targeting online advertisements.[205][206]

With social media sites overtaking TV as a source for news for young people and news organisations increasingly reliant on social media platforms for generating distribution,[207] major publishers now use artificial intelligence (AI) technology to post stories more effectively and generate higher volumes of traffic.[208]

There are a number of competitions and prizes to promote research in artificial intelligence. The main areas promoted are: general machine intelligence, conversational behavior, data-mining, robotic cars, robot soccer and games.

Artificial intelligence is breaking into the healthcare industry by assisting doctors. According to Bloomberg Technology, Microsoft has developed AI to help doctors find the right treatments for cancer.[209] There is a great amount of research and drugs developed relating to cancer. In detail, there are more than 800 medicines and vaccines to treat cancer. This negatively affects the doctors, because there are way too many options to choose from, making it more difficult to choose the right drugs for the patients. Microsoft is working on a project to develop a machine called "Hanover". Its goal is to memorize all the papers necessary to cancer and help predict which combinations of drugs will be most effective for each patient. One project that is being worked on at the moment is fighting myeloid leukemia, a fatal cancer where the treatment has not improved in decades. Another study was reported to have found that artificial intelligence was as good as trained doctors in identifying skin cancers.[210] Another study is using artificial intelligence to try and monitor multiple high-risk patients, and this is done by asking each patient numerous questions based on data acquired from live doctor to patient interactions.[211]

According to CNN, there was a recent study by surgeons at the Children's National Medical Center in Washington which successfully demonstrated surgery with an autonomous robot. The team supervised the robot while it performed soft-tissue surgery, stitching together a pig's bowel during open surgery, and doing so better than a human surgeon, the team claimed.[212]

Advancements in AI have contributed to the growth of the automotive industry through the creation and evolution of self-driving vehicles. As of 2016, there are over 30 companies utilizing AI into the creation of driverless cars. A few companies involved with AI include Tesla, Google, and Apple.[213]

Many components contribute to the functioning of self-driving cars. These vehicles incorporate systems such as braking, lane changing, collision prevention, navigation and mapping. Together, these systems, as well as high performance computers are integrated into one complex vehicle.[214]

One main factor that influences the ability for a driver-less car to function is mapping. In general, the vehicle would be pre-programmed with a map of the area being driven. This map would include data on the approximations of street light and curb heights in order for the vehicle to be aware of its surroundings. However, Google has been working on an algorithm with the purpose of eliminating the need for pre-programmed maps and instead, creating a device that would be able to adjust to a variety of new surroundings.[215] Some self-driving cars are not equipped with steering wheels or brakes, so there has also been research focused on creating an algorithm that is capable of maintaining a safe environment for the passengers in the vehicle through awareness of speed and driving conditions.[216]

Financial institutions have long used artificial neural network systems to detect charges or claims outside of the norm, flagging these for human investigation.

Use of AI in banking can be tracked back to 1987 when Security Pacific National Bank in USA set-up a Fraud Prevention Task force to counter the unauthorised use of debit cards. Apps like Kasisito and Moneystream are using AI in financial services

Banks use artificial intelligence systems to organize operations, maintain book-keeping, invest in stocks, and manage properties. AI can react to changes overnight or when business is not taking place.[217] In August 2001, robots beat humans in a simulated financial trading competition.[218]

AI has also reduced fraud and crime by monitoring behavioral patterns of users for any changes or anomalies.[219]

A platform (or "computing platform") is defined as "some sort of hardware architecture or software framework (including application frameworks), that allows software to run". As Rodney Brooks pointed out many years ago, it is not just the artificial intelligence software that defines the AI features of the platform, but rather the actual platform itself that affects the AI that results, i.e., there needs to be work in AI problems on real-world platforms rather than in isolation.

A wide variety of platforms has allowed different aspects of AI to develop, ranging from expert systems such as Cyc to deep-learning frameworks to robot platforms such as the Roomba with open interface.[221] Recent advances in deep artificial neural networks and distributed computing have led to a proliferation of software libraries, including Deeplearning4j, TensorFlow, Theano and Torch.

Amazon, Google, Facebook, IBM, and Microsoft have established a non-profit partnership to formulate best practices on artificial intelligence technologies, advance the public's understanding, and to serve as a platform about artificial intelligence.[222] They stated: "This partnership on AI will conduct research, organize discussions, provide thought leadership, consult with relevant third parties, respond to questions from the public and media, and create educational material that advance the understanding of AI technologies including machine perception, learning, and automated reasoning."[222] Apple joined other tech companies as a founding member of the Partnership on AI in January 2017. The corporate members will make financial and research contributions to the group, while engaging with the scientific community to bring academics onto the board.[223]

There are three philosophical questions related to AI:

Can a machine be intelligent? Can it "think"?

Widespread use of artificial intelligence could have unintended consequences that are dangerous or undesirable. Scientists from the Future of Life Institute, among others, described some short-term research goals to be how AI influences the economy, the laws and ethics that are involved with AI and how to minimize AI security risks. In the long-term, the scientists have proposed to continue optimizing function while minimizing possible security risks that come along with new technologies.[233]

Machines with intelligence have the potential to use their intelligence to make ethical decisions. Research in this area includes "machine ethics", "artificial moral agents", and the study of "malevolent vs. friendly AI".

The development of full artificial intelligence could spell the end of the human race. Once humans develop artificial intelligence, it will take off on its own and redesign itself at an ever-increasing rate. Humans, who are limited by slow biological evolution, couldn't compete and would be superseded.

A common concern about the development of artificial intelligence is the potential threat it could pose to mankind. This concern has recently gained attention after mentions by celebrities including Stephen Hawking, Bill Gates,[235] and Elon Musk.[236] A group of prominent tech titans including Peter Thiel, Amazon Web Services and Musk have committed $1billion to OpenAI a nonprofit company aimed at championing responsible AI development.[237] The opinion of experts within the field of artificial intelligence is mixed, with sizable fractions both concerned and unconcerned by risk from eventual superhumanly-capable AI.[238]

In his book Superintelligence, Nick Bostrom provides an argument that artificial intelligence will pose a threat to mankind. He argues that sufficiently intelligent AI, if it chooses actions based on achieving some goal, will exhibit convergent behavior such as acquiring resources or protecting itself from being shut down. If this AI's goals do not reflect humanity's - one example is an AI told to compute as many digits of pi as possible - it might harm humanity in order to acquire more resources or prevent itself from being shut down, ultimately to better achieve its goal.

For this danger to be realized, the hypothetical AI would have to overpower or out-think all of humanity, which a minority of experts argue is a possibility far enough in the future to not be worth researching.[239][240] Other counterarguments revolve around humans being either intrinsically or convergently valuable from the perspective of an artificial intelligence.[241]

Concern over risk from artificial intelligence has led to some high-profile donations and investments. In January 2015, Elon Musk donated ten million dollars to the Future of Life Institute to fund research on understanding AI decision making. The goal of the institute is to "grow wisdom with which we manage" the growing power of technology. Musk also funds companies developing artificial intelligence such as Google DeepMind and Vicarious to "just keep an eye on what's going on with artificial intelligence.[242] I think there is potentially a dangerous outcome there."[243][244]

Development of militarized artificial intelligence is a related concern. Currently, 50+ countries are researching battlefield robots, including the United States, China, Russia, and the United Kingdom. Many people concerned about risk from superintelligent AI also want to limit the use of artificial soldiers.[245]

Joseph Weizenbaum wrote that AI applications can not, by definition, successfully simulate genuine human empathy and that the use of AI technology in fields such as customer service or psychotherapy[246] was deeply misguided. Weizenbaum was also bothered that AI researchers (and some philosophers) were willing to view the human mind as nothing more than a computer program (a position now known as computationalism). To Weizenbaum these points suggest that AI research devalues human life.[247]

Martin Ford, author of The Lights in the Tunnel: Automation, Accelerating Technology and the Economy of the Future, and others argue that specialized artificial intelligence applications, robotics and other forms of automation will ultimately result in significant unemployment as machines begin to match and exceed the capability of workers to perform most routine and repetitive jobs. Ford predicts that many knowledge-based occupationsand in particular entry level jobswill be increasingly susceptible to automation via expert systems, machine learning[249] and other AI-enhanced applications. AI-based applications may also be used to amplify the capabilities of low-wage offshore workers, making it more feasible to outsource knowledge work.[250][pageneeded]

This raises the issue of how ethically the machine should behave towards both humans and other AI agents. This issue was addressed by Wendell Wallach in his book titled Moral Machines in which he introduced the concept of artificial moral agents (AMA).[251] For Wallach, AMAs have become a part of the research landscape of artificial intelligence as guided by its two central questions which he identifies as "Does Humanity Want Computers Making Moral Decisions"[252] and "Can (Ro)bots Really Be Moral".[253] For Wallach the question is not centered on the issue of whether machines can demonstrate the equivalent of moral behavior in contrast to the constraints which society may place on the development of AMAs.[254]

More:

Artificial intelligence - Wikipedia

South China Morning Post

Tencent Hires Baidu's Big-Data Director, Betting on Artificial Intelligence Wall Street Journal (subscription) SHANGHAIIn a sign of its ambition to compete more in artificial intelligence, Tencent Holdings Ltd. says it has hired Zhang Tong, a Stanford-trained researcher who formerly led Baidu Inc.'s Big Data Lab, to oversee its fledgling AI program. Dr. Zhang ... Tencent names Zhang Tong head of its artificial intelligence labSouth China Morning Post Tencent Joins Rush Into AI to Keep Lead in Social Media, GamingBloomberg all 56 news articles »

Read the original:

Tencent Hires Baidu's Big-Data Director, Betting on Artificial Intelligence - Wall Street Journal (subscription)

A couple of weeks ago, Coca-Colas global senior digital director Mariano Bosaztold Adweek he wanted to start experimenting with automated narratives, including using bots for music and editing the closingcredits of commercials.

Algorithms are already foundational to programmatic advertisingand will likely only grow to be a bigger part of media buying, but can machine learning ever completely replace the creative process? Its no surprise that agencies adamantly say no, that brands still need human creatives to handle strategy and come up with ideas. But creative shops are stillpreparingfor a timewhen there will be fewerpeople to handle some parts of the business, especially those that involve time-consuming and manual tasks.

To be honest, some of the first people who will lose their job because of AI will be marketing managers, said Firstborns executive creative director Dave Snyder. If your job is really to move numbers around a spreadsheet and optimizing it based on whats performing, the computer is going to be way better than you and faster.

Still, Andy Hood, head of emerging technologies for AKQA, said that shop has invested heavily in AI services but that the best is brought out of the human in the AI. At least for now.

At some point in the future, complete automation may be possibleit may even be desirable, Hood said. But I think for the more foreseeable future, were looking at these intelligent tools that combine with creative teams to find the best results.

Heres a look at how four agencies are using AI as part of the creative process:

AKQA is testing an internal tool it built using IBM Watson for clients that scours online platforms to find new groups of consumers for brands.

It finds new audiences and reaches out proactively to those new audiences that clients arent necessarily talking directly to, said AKQAs Hood. He declined to say which clients are using it, but you could imagine a travel brand being able to find peoplewho are talking about traveling without mentioning a specific brand.

However, the system is not totally hands-off. Hood referencedMicrosofts Taychatbot that spit out racist and anti-Semitic languageas an example of why AI still needs humans behind it.

It takes a degree of confidence in your automated system to just give it access to your public and turn it looseIm not quite sure that were there yet, Hoodsaid. There are still people involved in the process, but weve moved from just using the data and the machine learning to create testing and incredible targeting and actually brought it into the narrative and storytelling.

The shop is working with programmatic-creative platform Thunder to change digital ad creative on the fly, particularly regional and pricing information.

JWT Canada plans to use AI for its airline and bank clients to dynamically change parts of adslike someones location or the price of a flightwhile still using one idea across video, display and audio ads.

"Ultimately it leads to leaner, tighter briefs and work that can really move the needle but also be brand-building if you pull the right pieces together."

- Andrew Rusk, business director of demand creation at JWT Canada

[Targeting] specific markets is something that weve been able to do for decades, said Andrew Rusk, business director of demand creation at JWT Canada. Whats different is using data signals in order to figure out if youre talking to a traveler who is trying to go to Europe as opposed to someone who is traveling to Asia versus someone who has a family or is a single traveler.

The partnership showshow creative agencies are increasingly taking on media agency-like work such as targeting, noted Thunder CEO Victor Wong. We see creative agencies becoming more like consumer-experience agencies where theyre responsible for designing all the touch points for a brand, which oftentimes is paid media, Wong said. Theres still a person setting the strategy and vision, but the execution in the long term can be 100 percent robot.

That may sound scary tocreative shops, said Rusk, but ultimately, it leads to leaner, tighter briefs and work that can really move the needle but also be brand-building if you pull the right pieces together.

Firstborns Snyder said brands arent asking about AI explicitly. He thinks low-level collateral like regional car commercials with creative that typically gets slightly tweaked by location is ripe for algorithmic creative. I think a lot of that stuff 100 percent will be created through machine learning and AI, he said. Youre already seeing it exist nowload the system with a bunch of cuts, and then based on demographic data, it will kind of auto-create an edit thats appropriate for the demographic or psychographic.

He saidits likely to replace marketing manager positions or music editors, for example, who manually refine the sound of a commercial because there will be a point where you dont need that there, I bet.

But Snyderdisagreed with the notionthat AI can replace a creative idea or a human. It still requires training a system to work, he said. In a way, AI would just create a ton of more generic-feeling shotsits not really a narrative.

Saatchi & Saatchi LA has run three or four AI campaigns for brands including a Facebook campaign with Toyota in January that used 1,000 different interests to help people discover unusual activities using IBM Watson.

One ad, for instance, matched peoplewho had an interest in both martial arts and barbecueto serve ads encouraging them to try out an activity called taikwan tenderizer in whichthey used hand-to-hand combat to tenderize pieces of meat in their backyards.

It allows us to deep dive into insights that we normally wouldnt have access to, saidChris Pierantozzi, ecd of Saatchi LA. Were looking at machines to help us break into unique patterns that people have been talking about, behaviors that theyre doing or other things that we normally wouldnt see.

Pierantozzi calls those types of ads flexible storytelling, pieces within ads that can be changed based on data. You still have a story, you still have an idea, and what weve done is weve made the kind of stuff that makes it feel like its more personal to who you are, he said.

Read the rest here:

How 4 Agencies Are Using Artificial Intelligence as Part of the Creative Process - Adweek

LAS VEGAS: Adobe and Microsoft are jointly working on artificial intelligence (AI) to offer better products and provide customers more automated, intelligence-based experiences, a top Adobe official said here.

Brad Rencher, Executive Vice President and General Manager, Marketing, of Adobe said the two tech giants were working on standard data models and sharing of core libraries between Adobe's Sensei and Microsoft's Cortana, both based on artificial intelligence.

Cortana is a search tool which can verbally provide answers to search queries and Sensei - a set of intelligent services from Adobe - integrates the advertising, marketing and analytics products offered on Cloud with back up of creatives and documentation.

Rencher, who was talking to a group of journalists here at the Adobe's annual Summit, said the joint research and development would combine the specific domain capabilities of Sensei with the wider core data platform of Cortana, thus building a service.

Adobe products can now use data from Microsoft Dynamics 365, Microsoft Power BI and Microsoft Azure into Sensei for intelligent machine learning.

Sensei will soon enter into Microsoft tools.

Rencher, however, said no discussion had taken place on how to monetise the collaboration.

Talking of Adobe's presence in India, Rencher said that it was the fastest growing market and they have had a very substantial amount of the company's research taking place in India, including on Sensei.

Rencher said large Indian companies are rapidly adopting Adobe's products and Cloud offerings.

"Reliance Industries was looking at how to integrate data across all its various divisions and Adobe had helped a very old newspaper, Malayala Manorama, to completely digitise its functions across the board," noted Rencher.

Despite the enormous amount of research taking place on AI, he said he did not believe that it could replace the creative side of human beings.

"What AI can do is reduce the time taken in intelligent data crunching and sometimes understanding what went wrong very quickly," Rencher added.

"By cutting six months of manual research to, say two minutes, it adds huge amount of strength to the creative aspects of human beings," he noted.

See the original post here:

Adobe, Microsoft working together on artificial intelligence - Economic Times

Forbes

Why Commerce Players Must Invest In Artificial Intelligence Today Forbes There is no way to spell retail without AI. The power of artificial intelligence (AI) to transform the consumer journey dominated the conversation at Shoptalk this week. In its second year, Shoptalk brought together 5,000-plus players across the full ...

Read the rest here:

Why Commerce Players Must Invest In Artificial Intelligence Today - Forbes

In a changing competitive landscape where skilled nursing facilities, home health, assisted living and long-term care organizations are looking for ways to differentiate themselves, drive better care, and reduce the number of adverse events within their patient populations, healthcare analytics has a big role to play. But one expert said that analytics is only one piece of a much larger puzzle that is required to get the post-acute care field in order.

Analytics is an old-fashioned term for what folks are doing today, said Terry Sullivan, MD, chief medical information officer at Onpointe, a post-acute care provider organization. There is analytics, there is artificial intelligence, there is Big Data, and there is machine learning. They all come together in cognitive machines and cognitive clinical capabilities. Its the summation of those four things.

Sullivan said that if a post-acute care provider organization does not have analytics, is not measuring something, and is not using an electronic health record, it should close the door and leave town.

The world is moving faster than that, he said. In using analytics and artificial intelligence and these other pieces, they are demonstrating where they are at, which is crucial for them to be 21st century players in a competitive marketplace.

Learn more at theBig Data & Healthcare Analytics Forumin San Franciso, May 15-16, 2017. Register here.

These technologies analytics, AI, Big Data and machine learning enable providers to optimize the patient journey so they and their partners are capable of taking on financial risk, Sullivan explained.

At the end of the day, if you are not capable of taking risk and performing in an optimum way, then you are not going to be very competitive, he said. The technology is part of it. One part is you align with the right financial partners, so you are in either a risk situation, a bundle situation, you are aligned with somebody so you are not in a fee-for-service model. Then you have to have an operational model that connects to the people above you who are referring to you and below you who you are referring to. And you have to have an IT capability that allows you to optimize the patient journey.

Healthcare organizations require all of those things to perform, because without high-quality information an organization cannot really participate in the continuum of care and will not be competitive, Sullivan added.

At a minimum the information would involve an EHR that has connections to the referring group above you and to the group below you who you are referring to, he said.

One of the four components of Sullivans technology requirements, artificial intelligence, can help post-acute care organizations drive market differentiation and care quality, he said.

Today, patient care is driven by usual and customary practice; things like length of stay, place of stay, are driven by historical practice on average, people stay a certain amount of time at a certain location, he explained. With artificial intelligence and machine learning, then onto cognitive machines, you for the first time are becoming truly patient-centered not facility-centered. In the old days, I keep patients two or three weeks. In the new paradigm, I keep the patient there only until they can safely move to the next level of care because cognitive machines allow me to know more information about the patient that allows me to make that decision when they can safely move to the lower level and less expensive level of care.

Sullivan will discuss his experience using these technologies at the HIMSS and Healthcare IT News Big Data & Healthcare Analytics Forum, May 15-16, 2017, in San Francisco, during a session entitled Post-Acute Care: When Productions Move Outside the Hospital.

Twitter:@SiwickiHealthIT Email the writer: bill.siwicki@himssmedia.com

Like Healthcare IT News on Facebook and LinkedIn

See the original post:

CMIO says analytics, artificial intelligence, Big Data and machine learning are becoming IT requirements - Healthcare IT News

Uber has been in the news for all the wrong reasons of late. Among its many sins, he company's canterkours CEO Travis Kalanick recently argued with a driver over fares.

If you think that Kalanick doesn't really value the company's "driver-partners", you are correct. Uber may be exceptional in its CEOs acerbic temperament, but at least it has plenty of company in another regard: Uber is among many firms working on removing drivers from the equation altogether. Make no mistake: driverless cars are coming.

To this end, I recently sat down withVivek Wadhwa, author ofThe Driver in the Driverless Car: How Our Technology Choices Will Create the Future. (Wadhwa worked withformer BusinessWeek technology editor Alex Salkever.) The following are excerpts from our conversation.

PS: What was your motivation for writing the book?

VW: Advancing technologies such as artificial intelligence, robots, sensors, and synthetic biology have made amazing things possible. It has literally become possible to solve the grand challenges of humanityproblems such as disease, hunger, energy, and education. We could, within two or three decades, be in an era of abundance, in which we live long and healthy lives, have unlimited clean energy and education, and have our most basic wants and needs met. At the same time, these technologies are making is possible to create new nightmares: killer viruses, safety and security hazards, and a jobless future. And they are widening the gap between the haves and have nots. The people who are being left out are getting angry and resentful. We are seeing this anger manifest itself in many waysincluding the elections.

I wrote the book to start educating parents, teachers, students, policy makers, and thinkers about the choices we must make to ensure that we head into a utopian future rather than continue on the path of dystopia that we have taken. I honestly see the possibility of creating a Star Trek future300 years ahead of schedule. I also see the possibilities of Mad Max. And I am convinced that we can get this rightif we all learn the possibilities and make sensible choices.

PS: Why have companies made so much progress with driverless and semi-autonomous cars so recently?

VW: Artificial Intelligence, which we had written off as dead is alive and kicking. It is now advancing on an exponential curve. Advances in a new form of AI called machine learning, are allowing the software to program itself. The computer is taught what to learn and how to learn and makes its own decisions. These are modeling the human mind itself using techniques similar to our learning processes. Before, it could take millions of lines of computer code to perform tasks such image recognitionwhich is needed for self driving cars. Now it can be done in hundreds of lines. All that is required is a large number of examples so that the computer can teach itself.

PS: Ive read reports about how traffic in cities may drop considerably. New apartments may replace parking structures. Are there any other potential benefits beyond being able to work in your car while it drives you to work?

VW: We will be far more productive, distance will no longer be a barrierso we can live 150 miles away and still get to work in time, and accidents will largely be a thing of the past. Also the disabled will no longer struggle to find transportation, mothers will be able to send their children to school without worrying about them reaching safely, and everyone will be able to afford to be mobilebecause of the reliability, safety, and lower costs of these technologies. Also, women and children will never worry about getting a cab ride late at night. Driving While Black will no longer be a vicious form of discrimination.

PS: Weve seen taxicab companies violently resist Uber and Lyft. Truckers cant be happy either. Is disruption here inevitable? Why wouldnt they take to Washington?

VW: This is just the beginning. As industries get disrupted, we will have large scale social disorder. People will feel disenfranchised and lose the sense of identity that comes from work. We just arent ready for the changes that are about to happen. I have discussed the core technologies in Driver in the Driverless Car, but am now working on my next book about how we can deal with the jobless future that is inevitable. How will we manage the transition, what sort of policies do we need to develop, and what will the new social order look likethese are some of the things I am now researching.

PS: Is Ubers CEO intent on self-immolation?

VW: Travis Kalanick represents everything that is bad about the technology industry: the greed, sexism, arrogance, and lack of ethics and values. This is why I teach: to show my engineering students the good they can do with technology and to inspire them to use it to uplift humanity rather than destroy it by becoming like Kalanick and others who lack a sense of social purpose.

This is how I conclude my book. It summarizes in a different way, the questions you are asking:

I showed you a broad range of technologies and asked you to view them through a lens or filter to assess their value to society and mankind. I asked you to consider whether they had the potential to benefit everyone equally, what the risks and the rewards were, and whether the technology more strongly promote autonomy or dependence. It is fairness and equality that are at the heart of these questions. Industry disruption is going to happen; tens of millions of jobs are going to disappear; our lives will change for the better and for the worse. If we manage it equitably and ease the transition and pain for the people who are most affected and least prepared, we can get to Star Trek: we can be living in an era in which every one of us has food, shelter, education, and light, and is connected to all. We will have better lives if we can adapt quickly enoughpsychologically, socially, ethically, and legallyand adjust to a world that changes literally before our eyes, every day, every minute.

At the end of the day, I believe that you will figure this out and we, as a collective race, will figure this out. Despite my fears, I know that humanity will rise to the occasion and uplift itself because it always has. We wouldnt have gotten this far if we did not have it in us to rise to great occasions.

Of the Star Trek future, Captain Picard once said: The acquisition of wealth is no longer the driving force in our lives. We work to better ourselves and the rest of Humanity. That is the future that we must build together.

See original here:

Artificial Intelligence and Driverless Cars: An Interview with Vivek Wadhwa - Huffington Post

How Artificial Intelligence Helped to Create a Gaud-Inspired Thinking Sculpture

IBM and New-York-based design studio SOFTlab have teamed up to create the first thinking sculpture, inspired by Gaud and developed with IBMs Watsoncognitive technology for the Mobile World Congress in Barcelona, Spain.

In order to help design the sculpture, Watson was taught about the history and style of Gaud and the architecture of Barcelona through volumes of images, literary works, articles, and even music. From these references, Watson helped to uncover critical insights on patterns in Gaud's worklike crabs, spiders, and color palettesthat the design team didn't initially associate with Gaud. The resulting four-meter-tall sculpture features a structural surface made of over 1200 unique aluminum parts, and is unmistakably reminiscent of Gauds work both in look and feel, yet entirely distinct.

The sculpture was on display from February 27 to March 2 at Mobile World Congress in Barcelona, where it interacted with visitors by changing shape in real-time, in response to sentiments from Twitter. To learn more about the sculpture, ArchDaily was given to opportunity to speak with IBM Watson Manager Jonas Nwuke.

+10

Sabrina Santos: Why was Gaudis work in particular chosen to inspire the project?

Jonas Nwuke: Architects at SOFTlab knew Mobile World Congress 2017 would be hosted in Barcelona, which sparked them to create something they had never done before. The SOFTlab team worked with Watson through the inspiration of legendary Barcelona architect Antoni Gaudi to create a sculpture alive with data. Antoni Gaudi was an iconic architect that shaped the city of Barcelona with his avant-garde architecture that was light years ahead of its time. In turn, the city of Barcelona heavily shaped the work he created. SOFTlab wanted to work with Watson to bring this approach from the past to the present.

SS: Approximately what percentage of the project decisions were made by Watson, as opposed to IBM and SOFTlab designers?

JN: Watson acted as a guide throughout the design process, uncovering insights from Gaudis work to inspire the architects, rather than making decisions on its own. Watsons ability to process volumes of information from images and documents sparked new ideas and helped the architects to reimagine their construction. Our vision for Watson has always been to use cognitive computing to augment, rather than replace, human intelligence.

SS: Besides the crab and spider references, were there any other particularly interesting ideas or connections that Watson created?

JN: Watsons analysis of thousands of Gaudi-inspired works helped the team pick unique, transformative colorsultramarine blue, jade green, yellow and orangewhich in turn helped SOFTlab select the iridescent dichroic film that brings the sculpture to life. Gaudis work brings about clear themes like waves, undulations and arches. With Watson, the designers were able to see themes that werent as obvious previously, such as candy and shells, in addition to crabs and spiders. These elements helped inspire the hanging chains and funnels of the sculptures design.

SS: How long did it take for Watson to review the images, documents, colors, and so forth?

JN: The entire project took a little over one month to complete.

SS: Were any of these tools (images, documents, colors, etc.) more influential than others, or do the various mediums integrate evenly as inspiration?

JN: Watson reviewed thousands of images, literary works, articles and even music surrounding Gaudi and Barcelona. With this, Watson was able to integrate these mediums evenly to become a Gaudi expert that could help the team understand a century of inspiration to re-inspire the design process.

SS: Who is creating the sculpture this week in Barcelona? How long will this take?

JN: Over a dozen SOFTlab designers worked over one week to install the sculpture at Mobile World Congress in Barcelona.

SS:What are the measurements and materials of the sculpture?

The height on the sculpture is 4 meters. The structural surface is made of over 1200 unique aluminum parts. This surface was clad in 3M Dichroic film which has an iridescent quality and was chosen based on suggestions from Watson and inspired by Gaudi's interest in similar colors. This material in combination with data controlled lighting produces a range of coloration and cast light. The Dichroic film was then clad in petal like laser cut aluminum panels that are inspired by some of the tiled forms used by Gaudi. The gravity driven cables underneath are made of 100 meters of ball chain.

SS: How will Tweet tones influence the height of the sculpture (for example, do positive Tweets make the sculpture extend lower?)?

JN: The sculpture is responsive to social sentiment and uses the IBM Watson Tone Analyzer API to identify and react to sentiment from Twitter. The tone and sentiment extracted from the tweets will be reflected in the sculpture, which is comprised of 3 funnels representing distinct topics. Each funnel is comprised of rings that will shift in height based on a given 'big 5' personality trait. One funnel contains one ring that addresses how open people are to Artificial Intelligence at Mobile World Congress. The Openness score for each tweet related to AI will drive the ring height. As the confidence Watson determines in peoples openness changes, the height of the ring changes as well. The second funnel contains three rings that react to Watsons confidence level in how open people are to the top 3 trending topics of the moment. The third funnel has 5 rings addresses the collective buzz of the event. Each ring represents a 'big 5' personality trait (Passion, Joy, Excitement, Curiosity & Encouragement) and confidence scores will drive ring height.

SS: How do you envision Artificial Intelligencelike Watsonbeing influential to the architectural world in the future?

JN: We envision Watson will serve as an assistant to other architects, acting as an extension of the creative process in the future. Cognitive technologies have the ability to uncover facts and answer questions, and can be applied to invent and explore new frontiers, such as architecture. For example, beyond architecture, Watson is helping professionals in many creative industries. At the 2016 Met Gala, IBM and Marchesa unveiled a cognitive dress worn by model Karolina Kurkova that served as an innovative collaboration with cognitive woven into every step of the creative process from concept, to R&D, from design and alteration to the finished product. In music, Grammy award-winning music producer Alex Da Kid collaborated with IBM Watson to inspire his breakout song as an artist, Not Easy. For this partnership, Watson analyzed the last five years of culture and music data to uncover new emotional insights to augment Alexs creative process. We believe Watson will continue to serve as a resource for creative professionals looking for inspiration.

Learn more about the project here.

News via IBM.

Here is the original post:

How Artificial Intelligence Helped to Create a Gaud-Inspired Thinking Sculpture - ArchDaily

Strategy matters in war. But behind every good strategy is good data. Take Korean War veteran and US Air Force officer John Boyd as an example.

He was tasked with analysing the outcome of dogfights aerial battles between fighter planes conducted at close range and come up with a way to save the lives of more American pilots. He did.

What Boyd created was a framework for decision-making that is known as the OODA loop. OODA refers to the recurring cycle of four actions: observe, orient, decide and act. He discovered that the pilots who came out of dogfights most successfully were those who had processed the loop as quickly and as often as possible.

Their experience of reacting to lots of different situations meant they could best adapt their battle strategies to unfolding events.

The theory of the OODA loop is central to the activities of Sentient, which is the worlds most funded artificial intelligence (AI) company

What Boyd was telling us was that surviving, learning and adapting were key to winning, says Antoine Blondeau, the co-founder and co-chairman of Sentient Technologies. Since then, that framework has been pervasive within the US armed forces. If you train, train and train, you go through that loop as often as possible.

The theory of the OODA loop is central to the activities of Sentient, which is the worlds most funded artificial intelligence (AI) company and the inventor of the technology behind Siri, Apples intelligent personal assistant.

When it builds a decision-making system, it aims to embed AI at every stage of the OODA loop. It does this by using deep-learning neural networks that have algorithms inspired by the function and structure of the human brain.

And as far as decision-making mechanisms go, there is no better example than the human brain. We humans are a winning design, Mr Blondeau points out. We are a species that has adapted very, very well to our environment to the point of domination.

In the space of a few milliseconds, artificial intelligence is in effect emulating an evolutionary process that took place over millions of years. But applying that approach at scale requires going through the OODA loop as often as possible, which in turn requires a huge amount of computing power.

The next stage of AIs evolution is the involvement of AI systems in the development of other AI systems

This explains why Sentients neural networks rely on the rapid mathematical calculations of 5,000 graphics processing units (GPUs). Most personal computers do not need more than one GPU.

Sentient is focused on developing commercial AI technology for three particular sectors: e-commerce, financial markets trading and insurance. Theres a lot of interest in implementing AI within the world of insurance, says Mr Blondeau, not only for risk prediction and developing better models of risk, but also for universal intelligence problems understanding the what and the how that make customers buy.

Last year Sentient launched Ascend, its own universal intelligence product. Ascend is essentially a self-evolving website that retailers can use to give their customers a personalised shopping experience. It tests different images, messaging and button placements to come up with the best combination for a particular shopper.

The result of this degree of personalisation is a striking uplift in sales, according to Mr Blondeau, who cites the example of a US mobile phone reseller that has already rolled out the technology. You can get a 40pc increase in lead generation and a 30pc increase in conversions within a few weeks. Thats the power of evolution within this loop, where the system is learning every minute.

The next stage of AIs evolution is the involvement of AI systems in the development of other AI systems. Needless to say, Sentient is already working on this technology, which Mr Blondeau predicts will help to dramatically reduce the cost of developing smart systems in future.

Yet he does note that there needs to be an ethical debate around regulating for augmenting but not replacing humans.

You can quickly evolve to intelligent systems that are addressing complex problems quickly and efficiently, with a feedback loop that operates not just at the application level but at the network structure level as well, says Mr Blondeau. We call that neuro evolution. Its effectively a marriage of deep learners and evolution.

While the decision-making ability of AI has the potential to be of huge benefit to humans, relieving us of the burden of having to make certain decisions ourselves, Mr Blondeau believes that aspect of the technology is widely misunderstood.

People think of AI as big data, he says. Then they think of it as insights and predictions. At Sentient, we tend to think of it as making decisions. Whatever data and insight humans gain every day is in order to make decisions. So true AI has to be able to make decisions not just look at data or predict trends. It has to be part of the decision-making process. Ideally AI is a decision-maker.

Read the rest here:

Artificial intelligence: the role of evolution in decision-making - Telegraph.co.uk