WIPO Technology Trends 2019: Artificial Intelligence

3y ago
66 Views
2 Downloads
6.28 MB
158 Pages
Last View : 12d ago
Last Download : 3m ago
Upload by : Farrah Jaffe
Transcription

WIPOTechnology Trends 2019ArtificialIntelligence

WIPOTechnology Trends 2019ArtificialIntelligence

The user is allowed to reproduce, distribute, adapt, translateand publicly perform this publication, including for commercial purposes, without explicit permission, provided that thecontent is accompanied by an acknowledgement that WIPOis the source and that it is clearly indicated if changes weremade to the original content.Suggested citation: WIPO (2019). WIPO Technology Trends2019: Artificial Intelligence. Geneva: World Intellectual tives should not carry anyofficial emblem or logo, unless they have been approved andvalidated by WIPO. Please contact us via the WIPO websiteto obtain permission.For any derivative work, please include the followingdisclaimer: “The Secretariat of WIPO assumes no liability orresponsibility with regard to the transformation or translationof the original content.”When content published by WIPO, such as images, graphics,trademarks or logos, is attributed to a third-party, the userof such content is solely responsible for clearing the rightswith the right holder(s).To view a copy of this license, please visit e designations employed and the presentation of materialthroughout this publication do not imply the expression of anyopinion whatsoever on the part of WIPO concerning the legalstatus of any country, territory or area or of its authorities,or concerning the delimitation of its frontiers or boundaries.This publication is not intended to reflect the views of theMember States or the WIPO Secretariat. WIPO, 2019First published 2019The mention of specific companies or products of manufacturers does not imply that they are endorsed or recommended by WIPO in preference to others of a similar naturethat are not mentioned.Photo credit: cover montage created with images by Margarita Lyr / iStock / Getty Images Plus and Daria Dombrovskaya / iStock / Getty Images PlusPrinted in SwitzerlandWorld Intellectual Property Organization34, chemin des Colombettes, P.O. Box 18CH-1211 Geneva 20, SwitzerlandISBN: 978-92-805-3007-0Attribution 3.0 IGO(CC BY 3.0 IGO)

Artificial intelligenceis a new digitalfrontier that willhave a profoundimpact on the world,transforming the waywe live and work.WIPO Director General, Francis Gurry

Preface71IntroductionForewordThe past, present and futureof AI: what research andinnovation trends can reveal;the data used in this reportand how it is analyzed; anda scheme for categorizingAI technologies.818 2 TrendsAbout thecontributors in artificialintelligence10Evolutionof AI patentapplicationsandscientificpublications3The historical developmentof AI innovation: analysisof trends in patents andscientific literature sincethe emergence of AI, andbreakdown by techniques,functional applications andapplication fields.38 Acknowledgments12Overall trends emergingfrom the data and analysis ofchanges over time, by regionand industry and the mostprominent entities.30Keyplayers inAI patenting4 ExecutivesummaryThe top applicants for AIpatents: how companies anduniversities/public researchorganizations compare, whichentities are most active ineach area and where theyare filing.1358

Geography 7 Key issues Selected AIof patentarising from categoriesfilingsAI and policy and termsresponses146Comprehensive analysis of5120Markettrends8 The futurerelated to AI of AI and theAnalysis of the use of AIIP systemtechnologies: data on6acquisitions, funding, opensource and patent litigationand oppositions.104Furtherreading150 The opportunities presentedby AI for business, societyand culture, and how AI and IPrights interact with each otherin the promotion of innovation.138WIPO Technology Trends 201982 Views from AI expertson the key policy andregulatory issues raised by AIapplications, plus examples ofpolicies proposed or enactedin various jurisdictions.5geographical trends, based onthe offices where patents arefiled, including both first andsubsequent filings.

This new report aimsto shed light on thetrends in innovation inAI since the field firstdeveloped in the 1950s.WIPO Director General, Francis Gurry

PrefaceBased on a detailed study of patents and otherinformation, this new report aims to shed lighton the trends in innovation in AI since the fieldfirst developed in the 1950s. Research hasinvolved the analysis of patent data relatedto AI inventions, as well as data on scientificpublications, litigation filings and acquisitionactivity. These data findings are discussed indetail and accompanied by commentary andindustry perspectives from more than 20 of theworld’s leading experts in AI; more completedatasets, patent search methodology andexpert contributions are openly available onthe WIPO website. We hope that this report willprove an invaluable resource for businesses,researchers and policymakers in the field, aswell as for those general readers who want tofind out how AI research has so far developed –and where it is going.The analysis offers unique insights into trendsin AI techniques (i.e., the different approachesused in AI, such as machine learning andfuzzy logic), AI functional applications (suchas natural language processing, speechprocessing and computer vision) and AIapplication fields (i.e., those industries andother sectors in which AI innovation is beingThis report identifies the key players inAI from both the corporate and publicsectors across different research areas andindustries. Furthermore, the analysis of thedata and the inputs from AI experts addressmany of the policy issues raised by AI, suchas the regulation and control of data, theincentivization of further research, the role ofintellectual property (IP) protection and thedevelopment of human-centered and ethical AIto benefit all.Together, the analysis of technological trendspresented in this report and the voices of AIexperts collected here are a valuable newaddition to a growing knowledge base on AI.I hope that this contribution will help to shiftdebate away from speculative interpretationand toward evidence-based projections,thereby informing global policymaking onthe future of AI, its governance and the IPframework that supports it.Francis GURRYDirector GeneralWIPO Technology Trends 2019AI is fast becoming part of our everyday lives,changing how we work, shop, travel andinteract with each other. Yet we are only atthe beginning of discovering the many waysin which AI will have an impact on – andindeed challenge – business, society andculture. There are numerous misconceptionsand misgivings about the nature of AI, and inparticular the challenge it poses to humankind.Given these widely held reservations andconcerns, it is essential to have a factual basisfor policy discussions about innovation in AI.Through this report, we hope to contributeevidence and bring clarity to this importantarea of debate.put into practice). One of the most strikingfindings of the report is that 50 percent of all AIpatents have been published in just the last fiveyears – a remarkable illustration of how rapidlyinnovation is advancing in this field.7I am pleased to present the first report in a newflagship series, WIPO Technology Trends. Thisfirst edition features artificial intelligence (AI) asthe theme. It is a fitting topic for the launch ofthe series as AI is a cutting-edge technologywith impacts on a wide range of businessesand activities.

ForewordAndrew NgCEO, Landing AIand deeplearning.aineural network, the better the neural networkperformed. We started looking for the biggestcomputers we could find. That discovery, andthe idea of using distributed computing toscale up neural networks, led to the creation ofGoogle Brain in 2011. Since then, deep learninghas made great progress as we’ve collectedmore data and created the powerful computersto make it work.I remember watching the final chess matchbetween Deep Blue and Gary Kasparov whileI was a graduate student at MIT in the 1990s.When Kasparov resigned, I jumped out ofmy chair in excitement – the AI communityhad finally triumphed over the humanchess champion!As AI continues to develop and transformindustry after industry, I hope it will createmany more similarly joyful experiences. Wehave a duty to address serious issues relatingto unemployment and inequality, but we shouldalso never lose sight of the value and possibilitythat AI brings.Even without this “holy grail,” AI is alreadycreating massive economic value in the worldtoday. In covering AI, the media tends to focuson images, speech and natural languageprocessing because those types of dataare very human. Everyone can understandwhat it means for a computer to identify acat or recognize your speech. But a lot of theeconomic value today is driven through deeplearning on “structured data.” Think of this asmachine learning applied to massive Excelspreadsheets (only bigger than the biggestExcel spreadsheet in the world). This data – forexample, what trucks should be dispatchedwhen, what products to recommend to whatuser – is more specific to individual companiesand industries. Structured data gets lessattention because it’s not as visual or intuitive.But it is driving massive value today and willcontinue as incumbent companies across allindustries transform themselves with AI.This report illustrates some of thosepossibilities and where they are arising. One ofits findings is that deep learning is the biggestand fastest growing technique in AI. When Iwas at Stanford in the 2000s, my PhD student,Adam Coates, came into my office with achart showing that the more data you fed to aWith this potential to transform every industryand create so much economic value, AIpresents just as great a technological shift asthe Internet did. In the Internet era, companiesincluding Microsoft and Apple saw the Internettrend early, and their leaders made strategicdecisions to become true Internet companies.ForewordThis moment stands out for me, amongmany milestones in the development of AI,such as the rise of deep learning, lessons inscalability and the DARPA Grand Challenge.AI is automation on steroids, and we had“automated” the process of playing chess. Iwas once captain of my high school chessteam and used to play competitively, but atthat moment I was happy to retire and leave itto the computers. This was not a sad event forme; on the contrary, I was thrilled to give upchess this way.8But we still have a long way to go in the field.For example, a toddler can usually recognizea cat after just one encounter, but a computerstill needs more than one example to learn.We need to find ways to train computers ontraining datasets as small as 100, or even10. Manufacturing lines, for example, usecomputer vision to check for defects in partsbut hopefully they will never have a millionexamples of defects to train on. Effective“unsupervised learning” – learning withoutlabelled data – remains a holy grail of AI.

Every organization can leverage its data anddomain expertise to build pilot AI projects.They will then be able to build internal AIteams that push their industries forward andcreate unique AI solutions. I recently releasedthe AI Transformation Playbook to helporganizations in all industries navigate theirown AI transformations.In order to help more people and businessesharness the potential of AI, we should considerthe following three actions:1. Build more public–private partnerships:When companies and universities orgovernment organizations work together,both sides can benefit. For example,university researchers can access moredata and understand relevant problems,and corporations can understand the latestbreakthroughs in technology.2. Continue promoting the free and opensharing of AI knowledge and resources:Companies have joined universities asleading forces for publishing free andopen AI research. The arXiv repositoryhas made a huge difference to the AIrevolution, as has the hosting service GitHubbecause it dramatically accelerates thefree dissemination of ideas. Other fieldscan learn this lesson and move away frompaywalled journals.3. Promote increased understanding of AI:Today, you can get a PhD to learn AI, but it’snot the only option. Traditional degrees, jobsthat allow you to learn, and at-home onlineToday, a significant amount of AI researchand education is taking place in the UnitedStates and China. These countries are hometo many of the best universities in the world,and their governments have provided fundingand created thoughtful regulations that enableinnovation. But the United States and Chinahave also built incredible business ecosystems.It’s very difficult for other countries, even thosewith great education, to compete with thebusiness, engineering and investing talent ofthose two countries.To close this gap, and work towardsdecentralizing the concentration of AItechnology, we must invest in education.Governments should invest heavily in educatingtheir citizens and enter public–private partnerships to adopt AI-powered systems safely.AI will transform every facet of society. Itbrings tremendous promise to improve ourlives and the world we live in, but it will requirethe creation of an AI ecosystem to ensurelong-term, sustainable growth. As this reporthighlights, AI technology and innovation hasuntil now been focused on a small numberof regions and organizations. In building afairer and more equitable AI-driven society,we must empower businesses, governmentsand citizens who may be impacted byautomation to ensure that the benefits of AI arewidely shared.WIPO Technology Trends 2019We are seeing a small number of incumbentcompanies undergo a similar transformationin the AI era. Google and Baidu were leadersin hiring AI talent and building pilot AI projects;they are now fully AI-driven companies. But AItransformation is not limited solely to Internetsearch companies. The biggest untappedopportunities today lie outside of the softwareindustry in industries such as agriculture,healthcare and manufacturing.learning programs (such as MOOCs) areall good ways for people to learn and workon important problems. The more widelyaccessible information we have, the moreflexibility people will have to learn.9This meant investing in the right technology,building teams, developing a digital strategyand adapting how they communicated abouttheir products.

About the contributorsThis report includes contributions fromexperts in AI, data, intellectual property,policy, and innovation. Their viewpoints andcomments complement and add contextto the information revealed in patent data,addressing issues such as existing andpotential uses and impact of AI technology,legal and regulatory questions, dataprotection and ethical concerns.Seth G Benzell is postdoctoralassociate at the MassachusettsInstitute of Technology (MIT)Initiative on the Digital EconomyNick Bostrom is Director of theFuture of Humanity Institute andauthor of Superintelligence: Paths,Dangers, StrategiesErik Brynjolfsson is Director at theMIT Initiative on the Digital EconomyYoon Chae is a senior associateBaker McKenzieFrank Chen is a partner atAndreessen HorowitzMyriam Côté is Director of AI forHumanity at Montreal Institute forLearning Algorithms (Mila)Boi Faltings is Professor of AIand Director of the AI Lab atÉcole polytechnique fédérale deLausanne (EPFL)Kay Firth-Butterfield is Head of AIand Machine Learning at the WorldEconomic Forum (WEF)10About the contributorsJohn G. Flaim is a partner andglobal head of the IP group atBaker McKenzieDario Floreano is Director of theLaboratory of Intelligent Systemsat EPFL and founding directorof the Swiss National Center ofCompetence in Robotics

Jay Iorio is a futuristRosalind Picard is founderand Director of the AffectiveComputing Research Group atthe Massachusetts Institute ofTechnology (MIT) Media LaboratoryMalcolm Johnson is DeputySecretary General of theInternational TelecommunicationsUnion (ITU)Hefa Song is Professor ofthe Institute of Science andManagement, Deputy Director ofthe Center for IPR Research andTraining, and Deputy Dean of theIntellectual Property School at theChinese Academy of Sciences (CAS)Konstantinos Karachaliosis Managing Director ofIEEE Standards Associationand a member of IEEE’sManagement CouncilPetr Šrámek is the founder of AIStartup Incubator and co-founderof the Platform on AI at theConfederation of Industry of theCzech RepublicKai-Fu Lee is founder of SinovationVentures and author of AISuperpowers: China, Silicon Valley,and the New World OrderAristotelis Tsirigos is Directorof the Applied BioinformaticsLaboratories at NYU School ofMedicineBen Lorica is Chief Data Scientistat O’Reilly MediaMiguel Luengo-Oroz is Chief DataScientist at UN Global PulseKazuyuki Motohashi is Professorin the Department of TechnologyManagement for Innovation at theUniversity of TokyoPaul Nemitz is Principal Adviser tothe European CommissionHaifeng Wang is Senior VicePresident at BaiduHerbert Zech is Professor of LifeSciences Law and IntellectualProperty Law at the Universityof BaselWIPO Technology Trends 2019Martin Ford is a futurist and authorof Rise of the Robots: Technologyand the Threat of a Jobless FutureEleonore Pauwels isResearch Fellow on EmergingCybertechnologies at theUN University’s Center forPolicy Research11Dominique Foray is Professor ofEconomics at EPFL

AcknowledgmentsThe report was prepared under the directionof Francis Gurry (Director General) and YoTakagi (Assistant Director General, GlobalInfrastructure Sector), supervised by AlejandroRoca Campaña (Senior Director, Accessto Information and Knowledge Division),under the responsibility of Irene Kitsara (IPInformation Officer, Access to Information andKnowledge Division).The report draws on commissionedbackground research, based on searchstrategy and methodology developed by thecore analytics team led by Irene Kitsara (WIPO),consisting of Sophie Gojon, Adrien Migeon andPhilippe Petit (CNRS Innovation) and PatriceLopez (science-miner). The AI dimensions usedfor the report and the related glossary weredeveloped by Patrice Lopez, who also providedexpert advice on AI in patent literature, withinputs by the core team, the WIPO AdvancedTechnologies Application Center (ATAC) andteam members of Mila (Simon Blackburn,Pierre Luc Carrier, Mathieu Germain, MargauxLuck, Gaétan Marceau Caron and JoaoFelipe Santos).Interviews were conducted and AI expertcontributions were compiled by James Nurton,who was the editor of the report, under theresponsibility of Charlotte Beauchamp (Head,Editorial and Design Section). Case studieswere kindly provided by contributors, AngelaHarp (IBM Research), Mohamad Ali Mahfouz(Microsoft Switzerland) and Sven Zirnite(Siemens Healthcare).12AcknowledgmentsThe report team benefited greatly fromexternal reviews of the draft chapters byBoi Faltings (EPFL), Patrice Lopez (scienceminer) and Alexandros Tsirigos (NYU MedicalSchool), and by WIPO colleagues Carsten Fink(Chief Economist), Akshat Dewan and BrunoPouliquen (from ATAC), and Marco Aleman,Allison Mages and Tomoko Miyamoto (fromWIPO’s Patent Law Division).The core team was skillfully assisted bycolleagues from WIPO’s Access to Informationand Knowledge Division, under the directionof Andrew Czajkowski. Thanks are due inparticular to Alica Daly, who validated the data,prepared the visualizations and contributedtogether with Alex Riechel to reviewing andfinalizing the report. Vipin Saroha preparedthe maps. Julie Summers provided valuableadministrative support. Additional dataand comments for the comparison of thereport findings with overall patent statisticswere provided by Kyle Bergquist, MosahidKhan, Julio Raffo and Hao Zhou, all from theEconomics and Statistics Division.Gratitude is due to the 27 leading experts in AIand contributors to the report who shared theirviews, provided their comments on differentaspects which enriched and contextualized thereport findings – thei

Artificial intelligence is a new digital frontier that will have a profound impact on the world, transforming the way we live and work. WIPO Director General, Francis Gurry. Preface 7 Foreword 8 About the contributors 10 Acknowl- edgments 12 Executive summary 13 1 Introduction The past, present and future of AI: what research and innovation trends can reveal; the data used in this report and .

Related Documents:

The Experience of China, India, Japan, Philippines, the Republic of Korea, Singapore and Thailand For more information contact the World Intellectual Property Organization Address: 34, chemin des Colombettes P.O. Box 18 CH-1211 Geneva 20 Switzerland Telephone: 41 22 338 91 11 Fax: 41 22 733 54 28 e-mail: wipo.mail@wipo.int Visit the WIPO .

USING XML (EXTENSIBLE MARKUP LANGUAGE) Version 1.4 Revision approved by the Committee on WIPO Standards (CWS) at its eighth session on December 4, 2020 Editorial Note prepared by the International Bureau At its fifth session, the Committee on WIPO Standards (CWS) agreed that the transition from WIPO Standard ST.25 to Standard ST.26 takes place in January 2022. Meanwhile, Standard ST.25 should .

2006 – 2014 Independent Consultant on international patent filing issues, specializing in the use of the PCT World Intellectual Property Organization (WIPO) Instruction of WIPO PCT seminars (over 90 to date). Staffing of a WIPO PCT Help Desk, answering questions posed by US and Canadian practitioners on various issues related to the

Legal text index History of the PCT Regulations . the World Intellectual Property Organization 1211 Geneva 20, Switzerland Telephone: 41 22 338 91 11 E-mail: wipo.mail@wipo.int Internet: www.wipo.int To order the PCT Newsletter please contact the elec-

3. The World Intellectual Property Organization (WIPO) The World Intellectual Property Organization (WIPO) is a specialized agency of the United Nations. It is dedicated to developing a balanced and accessible international intellectual property (IP) system, which rewards creativity, stimulates innovation and contributes to

WIPO Case No D BIZ 2002-00081 Baccarat SA v Priority Software, July 9th 2002. WIPO Case No D 2005-0048 Baccarat S.A v MSL International, March 7, 2005 WIPO Case No D 2006- 0038 Baccarat S.A. v Web Domain Names The Complainant therefore

(SMEs), women and youth in economic growth, they will be a special focus for WIPO's work. This is particularly true in terms of the need for IP skills-building and this focus is a new feature in the training courses and pro-grams offered in the WIPO Academy Portfolio of Education, Training and Skills Development Programs 2022.

Focus on programming, not the programming language. I included the minimum useful subset of Java and left out the rest. I needed a title, so on a whim I chose How to Think Like a Computer Scientist. My first version was rough, but it worked. Students did the reading, and they understood enough that I could spend class time on the hard topics, the interesting topics and (most important .