3y ago
396.83 KB
20 Pages
Last View : 5d ago
Last Download : 4m ago
Upload by : Joao Adcock


Since its founding in 1990, the McKinsey Global Institute (MGI) has soughtto develop a deeper understanding of the evolving global economy. As thebusiness and economics research arm of McKinsey & Company, MGI aimsto provide leaders in the commercial, public, and social sectors with the factsand insights on which to base management and policy decisions. For thesecond year running, the Lauder Institute at the University of Pennsylvaniaranked MGI the world’s number-one private-sector think tank in its 2016Global Think Tank Index.MGI research combines the disciplines of economics and management,employing the analytical tools of economics with the insights of businessleaders. Our “micro-to-macro” methodology examines microeconomicindustry trends to better understand the broad macroeconomic forcesaffecting business strategy and public policy. MGI’s in-depth reports havecovered more than 20 countries and 30 industries. Current research focuseson six themes: productivity and growth, natural resources, labor markets, theevolution of global financial markets, the economic impact of technology andinnovation, and urbanization.Recent reports have assessed the economic benefits of tackling genderinequality, a new era of global competition, Chinese innovation, anddigital globalization. MGI is led by four McKinsey & Company seniorpartners: Jacques Bughin, James Manyika, Jonathan Woetzel, andFrank Mattern, MGI’s chairman. Michael Chui, Susan Lund, Anu Madgavkar,Sree Ramaswamy, and Jaana Remes serve as MGI partners. Project teamsare led by the MGI partners and a group of senior fellows, and includeconsultants from McKinsey offices around the world. These teams draw onMcKinsey’s global network of partners and industry and management experts.Input is provided by the MGI Council, which coleads projects and providesguidance; members are Andres Cadena, Sandrine Devillard, Richard Dobbs,Katy George, Rajat Gupta, Eric Hazan, Eric Labaye, Acha Leke, Scott Nyquist,Gary Pinkus, Oliver Tonby, and Eckart Windhagen. In addition, leadingeconomists, including Nobel laureates, act as research advisers. The partnersof McKinsey fund MGI’s research; it is not commissioned by any business,government, or other institution. For further information about MGI and todownload reports, please visit McKinsey & Company 20172McKinsey Global Institute

IN BRIEFARTIFICIAL INTELLIGENCE:IMPLICATIONS FOR CHINAArtificial intelligence, or the idea that computer systems can perform functions typically associated withthe human mind, has gone from futuristic speculation to present-day reality. When the AlphaGo computerprogram defeated Lee Sedol, a nine-dan professional master, at the game of Go in 2016, it signaled to theworld that it is indeed possible for machines to think a bit like humans—and even exceed their capabilities.Thanks to advances in data collection and aggregation, algorithms, and processing power, computerscientists have achieved significant breakthroughs in artificial intelligence. Where computer systems oncehad to be programmed to execute rigidly defined tasks, they can now be given a generalized strategyfor learning, enabling them to adapt to new data inputs without being explicitly reprogrammed. Todaymany machine learning systems have already been developed for commercial use. The applicationsare tremendously varied, and adoption is growing rapidly in sectors such as finance, health care,and manufacturing.Because they can dramatically boost productivity, AI technologies may have a disruptive impact onChina’s economic growth and on its workforce. A McKinsey Global Institute study published earlier thisyear estimated that half of all work activities in China could be automated, making it the nation with theworld’s largest automation potential. Hundreds of millions of Chinese workers could be affected, and jobsmade up of routine work activities and predictable, programmable tasks will be particularly vulnerable.While impact on the labor market is likely to be gradual at the aggregate level, it can be sudden anddramatic at the level of specific work activities, rendering some jobs obsolete fairly quickly. Overall, AIwill raise the premium placed on digital skills while reducing demand for medium- and low-skill workers,potentially exacerbating income inequality.On the plus side, AI’s effect on productivity could be crucial to China’s future economic growth as thepopulation ages. According to a McKinsey Global Institute report, AI-led automation can give the Chineseeconomy a productivity injection that would add 0.8 to 1.4 percentage points to GDP growth annually,depending on the speed of adoption.With its biggest tech companies driving significant investments in R&D, China is one of the leading globalhubs of AI development. Its advantages include a vast population and diverse industry mix that havethe potential to generate huge volumes of data and provide an enormous market. But China will needto focus on building its innovation capacity. The United States and the United Kingdom are currentlyproducing more influential AI research, and the more robust US ecosystem nurtures more competitiveAI startups. Realizing AI’s economic potential in China also depends on its actual adoption—not justamong the technology giants but across China’s traditional industries. Achieving this goal will requirebuilding strategic awareness among business leaders, developing technical know-how, and overcomingimplementation costs.AI capabilities have exciting and far-reaching potential to enhance human welfare by improving healthcare, the environment, security, and education. At the same time, AI also raises complex ethical, legal, andsecurity questions surrounding issues such as privacy, discrimination, liability, and regulation. Prudentialgovernance should be put in place as AI is introduced into society on a broader scale.Although the market will drive the development and adoption of AI, the right policy framework canestablish a healthy environment for growth. Five priorities can form the basis of China’s AI strategy:building a robust data ecosystem; spurring adoption of AI within traditional industries; strengthening thepipeline of specialized AI talent; ensuring that education and training systems are up to the challenge; andestablishing an ethical and legal consensus among Chinese citizens and in the global community.The technology industry is becoming increasingly global. China has the capability and opportunity tolead international collaboration in the development and governance of AI, ensuring that this breakthroughtechnology will positively contribute to the general welfare of all humanity.

ARTIFICIAL INTELLIGENCE:IMPLICATIONS FOR CHINAArtificial intelligence (AI), or the idea that computer systems can perform functions typicallyassociated with the human mind, has suddenly gone from futuristic speculation to presentday reality. Computer scientists have achieved significant breakthroughs in machinelearning and deep learning, giving machines cognitive and predictive capabilities. Todaythese systems are already being deployed in real-world situations.A JOURNEY TO THE BRINK OF TRANSFORMATIONAI is defined as machines mimicking cognitive functions typically associated with the humanmind. This concept has long been the stuff of speculation and science fiction—and theoptimism surrounding it intensified after a number of initial theoretical advances were madein the 1950s and ’60s. But that wave of momentum fizzled in the face of technical obstacles.With AI failing to meet expectations, the field experienced a long fallow period.1 Subsequentdecades brought a few successes (such as IBM’s “Deep Blue” supercomputer defeatingGary Kasparov in chess), but the real-world use cases were too isolated to supportmass commercialization.Fast forward to the 21st century. Breakthroughs in data collection and aggregation,computing power, and algorithms (especially machine learning) enabled revolutionarytechnical advances. In one widely hailed milestone, Google’s AlphaGo computer programdefeated a human champion in the game of Go, which had been traditionally regarded asunsolvable by machines.But the advances are not only happening at the theoretical frontiers of the field. Analyticstools utilizing machine learning are the precursors of tomorrow’s super-intelligent systems,and many of them are already on the market. Adoption is growing rapidly in sectors suchas finance, health care, and manufacturing. Global venture capital funding has grown from 589 million in 2012 to over 5 billion in 2016.2 McKinsey estimates that the total market forAI applications will reach 127 billion by 2025.UNDERSTANDING AI AND WHAT IT CAN DOTraditionally, we have used the processing power of computers to generate output moreefficiently (for example, doing faster and more complex computations than humanscan perform). Conventional software programs have always been coded with specificinstructions on the tasks they need to execute.AI systems take a very different approach. They can sift through enormous “big data” setsto find patterns, associations, and insights—and as they do, they employ a generalizedstrategy for learning. This enables them to adapt to new data inputs without being explicitlyreprogrammed. Systems utilizing machine learning3 have induction and decision-makingcapabilities—and the systems being developed on the frontiers of this field push the1232Daniel Crevier, AI: The tumultuous history of the search for artificial intelligence, Basic Books, 1993.“The 2016 AI recap: Startups see record high in deals and funding,” CB Insights blog, January 19, telligence-startup-funding/.Machine learning is one of the most important technical developments in the field of AI. Building on thepremise that the human cognitive process can be represented by mathematical models, it feeds a hugevolume of data into an algorithm that is essentially a generalized strategy for learning. It then “trains” themachine to derive a rule or procedure for interpreting data or making predictions.McKinsey Global InstituteArtificial intelligence: Implications for China

boundaries even further with deep learning.4 These computer systems can learn, discover,and apply rules by themselves.While recent breakthroughs in deep learning have produced AI systems that can match orsurpass human intelligence in certain key functions, we are still decades away from “generalAI”—or machines that can perform the full range of cognitive tasks that humans can do. Butmany machine learning systems have already been developed for specific commercial uses,and the applications are tremendously varied. They can provide customer service, managelogistics, monitor equipment on factory floors, optimize energy consumption, and analyzemedical records. Recent McKinsey Global Institute (MGI) research finds that machinelearning techniques have wide applicability in virtually every industry.5It is useful to think of AI capabilities in four main categories: Perception involves collecting and interpreting information to sense the world anddescribe it. These capabilities include natural language processing, computer vision, andaudio processing. Prediction involves using reasoning to anticipate behaviors and results. Suchtechnologies are used, for example, to develop precisely targeted advertising forspecific customers. Prescription is principally concerned with what to do to achieve goals. It has a variety ofuse cases, including route planning, drug discovery, and dynamic pricing. Last but not least, AI can be combined with complementary technologies such asrobotics to provide integrated solutions. These include autonomous driving, roboticsurgery, and household robots that respond to stimuli.The current degree of commercialization varies for each type of AI functionality. Whilesystems with perceptive and predictive capabilities are already on the market, moreprescriptive tools and integrated solutions are still under development (Exhibit 1).AI’S FUTURE TRAJECTORY: PROFOUND CHALLENGES AND POSSIBILITIESThe technology advances of the past mainly enhanced capabilities to execute clearlydelineated tasks in production. But now AI enables machines to react and adapt in orderto optimize results. Together with technologies such as the Internet of Things (IoT) androbotics, it can create an integrated cyber-physical world.Current momentum points to the likelihood that AI technologies will eventually beembraced globally in an even wider variety of settings and industries—and one of the mostimportant consequences would be machines handling a variety of tasks that have alwaysbeen performed by humans. An MGI report analyzed more than 2,000 work activitiesacross 800 occupations in the global economy. Already it appears to be technicallyfeasible that 50 percent of today’s work activities could be automated using currentlydemonstrated technologies.But technical feasibility is only one factor affecting the pace and extent of automation.Others include the cost of developing and deploying specific applications, labor market45McKinsey Global InstituteDeep learning is a sub-branch of machine learning on the frontier of computer science. This technologyinvolves software-based calculators that approximate the function of neurons in a brain; they are connectedto form a hierarchical “neural network.” Instead of the “shallow” learning algorithms and manual featureextraction that characterize traditional machine learning, deep learning runs data inputs through multiplelayers of non-linear processing units, automatically extracts data features, and uses the previous layer’soutput as an input for the next layer. The intricacy of these neural networks enables even more sophisticatedcapabilities, such as image recognition and natural language generation.The age of analytics: Competing in a data-driven world, McKinsey Global Institute, December 2016.Artificial intelligence: Implications for China3

dynamics, the economic benefits, and regulatory and social acceptance. Taking thesefactors into account, MGI’s research on automation finds that it might take until 2055 forhalf of all current work activities to actually become automated—but there is a fair degree ofuncertainty in this timing. In an aggressive adoption scenario, that level of automation couldoccur 20 years sooner, and in a late adoption scenario, it might occur 20 years later.6Further down the road, AI could be a powerful tool to apply to some of society’s centralchallenges. In health care, AI will greatly enhance our capability to analyze the humangenome and develop personalized and more effective treatments for each patient. It couldradically accelerate efforts to cure cancer, Alzheimer’s, and other diseases. AI systems cananalyze weather patterns and improve energy efficiency on a wide scale, enhancing ourability to monitor and combat climate change. And the possibilities are not even earthbound;AI systems could one day pioneer exploration of Mars and the outer reaches of space.Exhibit 1The current degree of commercialization varies across AI technologiesFunctionality Maturity Use utionsIBMSoftware can review medical images with the same accuracy as humanradiologistsIflytekVoice assistant app transcribes spoken Mandarin into textNetflixAlgorithm suggests films and TV shows to customers based on their previousviewing history and ratingsCapital OneAlgorithm predicts customers’ purchasing behaviorWealthfrontAI-driven platform provides automated advice to customers on asset allocationand wealth managementGoogleAI can produce surrealistic “artwork” from white noise or imagesAmazonSmart speaker devices can control home appliancesBaiduAutonomous cars operate within known and limited environmentsSOURCE: MIT Technology Review; TechNode;; Google Research blog; McKinsey Global Institute analysisWHAT DOES AI MEAN FOR CHINA?With its biggest tech companies driving momentum for R&D, China is one of the leadingglobal hubs of AI development. Its vast population and diverse industry mix have thepotential to generate huge volumes of data and provide an enormous market. Wideadoption of AI technologies could be crucial to China’s future economic growth as thenation’s population ages, heightening the need to accelerate productivity growth. Some ofthe required building blocks include a more open data environment and well-trained datascience talent. But AI also poses complex social and economic questions that will requirecareful consideration.CHINA’S POSITION IN AI DEVELOPMENTChina and the United States are currently the world leaders in AI development. In 2015alone, they accounted for nearly 10,000 papers on AI published in academic journals, whilethe United Kingdom, India, Germany, and Japan combined to produce only about half asmany scholarly research articles.7674This assumes that the human labor replaced by automation would rejoin the workforce and be as productiveas it was in 2014. A future that works: Automation, employment, and productivity, McKinsey Global Institute,January 2017.SCImago Journal & Country Rank, 2015McKinsey Global InstituteChinaAIArtificial intelligence: Implications for China

Much of the momentum behind AI in China is being driven by private-sector techfirms. Aided by huge volumes of search data and their many product lines, some ofChina’s Internet giants are on the cutting edge of technologies such as image and voicerecognition.8 These capabilities have been integrated into new products, includingautomated personal assistants, autonomous cars, and so forth.China has reason to feel optimistic about its role in a future defined by AI. Its huge populationcan generate a tremendous volume of data, which is a prerequisite for “training” AI systems.China also has the advantage of “economies of scope”: its wide range of industries providea fertile market for deployment.But it will take a sustained effort to stay at the forefront of such a rapidly evolving field andmaximize the economic potential of these technologies. China will need to focus intently onbolstering its capacity for innovation. For example, while Chinese academics have actuallypublished even more papers on AI than US researchers, their papers have not generatedthe same impact as those by US or UK authors (Exhibit 2).Exhibit 2Although China produces a large number of widely cited AI-related papers, US and UK research remainsmore influentialWhile China ranks first for absolute AI citations,the United States holds an edge whenself-citations are taken outNumber of AI publications citedSelf-citations1Publication influenceH-index2Other itedStatesUnitedKingdomChinaGermany Canada1 Self-citation occurs when a journal cites another article published in the same journal.2 The H-index ranks both the productivity of scholars and the citation impact of their publications. A higher H-index number indicates more publications that arewidely cited.SOURCE: SCImago Journal Rank 2015; McKinsey Global Institute analysisFurthermore, China does not yet have the same kind of vibrant AI ecosystem as the UnitedStates, which has produced substantially more AI startup companies than China (Exhibit 3).The US ecosystem is large, innovative, and diverse (including research institutions anduniversities as well as private companies). Building on all the well-established strengths ofthe Silicon Valley tech sector, it has considerable advantages that are difficult to replicate.8McKinsey Global Institute“Why deep learning is sudde

Artificial intelligence, or the idea that computer systems can perform functions typically associated with the human mind, has gone from futuristic speculation to present-day reality. When the AlphaGo computer program defeated Lee Sedol, a nine-dan professional master, at the game of Go in 2016, it signaled to the world that it is indeed possible for machines to think a bit like humans—and .

Related Documents:

Bruksanvisning för bilstereo . Bruksanvisning for bilstereo . Instrukcja obsługi samochodowego odtwarzacza stereo . Operating Instructions for Car Stereo . 610-104 . SV . Bruksanvisning i original

Artificial Intelligence and Its Military Implications China Arms Control and Disarmament Association July 2019 What Is Artificial Intelligence? Artificial intelligence (AI) refers to the research and development of the theories, methods, technologies, and application systems for

Artificial Intelligence -a brief introduction Project Management and Artificial Intelligence -Beyond human imagination! November 2018 7 Artificial Intelligence Applications Artificial Intelligence is the ability of a system to perform tasks through intelligent deduction, when provided with an abstract set of information.

Artificial Intelligence and Its Implications on Future Submarine Warfare With reference to any historical example, what are the implications for the future of the Royal Australian Navy? This essay will discuss artificial intelligence (AI) a

10 tips och tricks för att lyckas med ert sap-projekt 20 SAPSANYTT 2/2015 De flesta projektledare känner säkert till Cobb’s paradox. Martin Cobb verkade som CIO för sekretariatet för Treasury Board of Canada 1995 då han ställde frågan

service i Norge och Finland drivs inom ramen för ett enskilt företag (NRK. 1 och Yleisradio), fin ns det i Sverige tre: Ett för tv (Sveriges Television , SVT ), ett för radio (Sveriges Radio , SR ) och ett för utbildnings program (Sveriges Utbildningsradio, UR, vilket till följd av sin begränsade storlek inte återfinns bland de 25 största

Hotell För hotell anges de tre klasserna A/B, C och D. Det betyder att den "normala" standarden C är acceptabel men att motiven för en högre standard är starka. Ljudklass C motsvarar de tidigare normkraven för hotell, ljudklass A/B motsvarar kraven för moderna hotell med hög standard och ljudklass D kan användas vid

LÄS NOGGRANT FÖLJANDE VILLKOR FÖR APPLE DEVELOPER PROGRAM LICENCE . Apple Developer Program License Agreement Syfte Du vill använda Apple-mjukvara (enligt definitionen nedan) för att utveckla en eller flera Applikationer (enligt definitionen nedan) för Apple-märkta produkter. . Applikationer som utvecklas för iOS-produkter, Apple .