みなさんこんにちは。フォーブスジャパンの谷本ゆかです。さて始まりましたフォーブスジャパンPWC AI Summit 2020。経営マネジメント確信にAIを組み込んだAI経営の最前線をDXの有識者、DXを実践する大企業のCDをデジタル部門責任者。そしてAI経営実現のための包括的支援サービスを提供するPWC JAPANのエキスパートときもとくイベントとなっています。まずは台湾デジタル担当大臣のオードリータンさんをゲストに迎えまして、AI FOR THE FUTURE、AIがもたらすビジネス変革と大事、PWC JAPAN合同会社、マネジングディレクター、エクスペリエンスセンターのまぶしクニオイスさんとの対談をお届けしてまいります。それではAI Summit 2020スタートです。今日はAi Summitのホストを紹介します。I'm superbiting a proposition to implement AI-driven management for large Japanese companies based on PLG's global AI capability. May I ask you some questions?Yes, of course.Thank you.I have that Taiwan has AI for the government and it's one of the most digital advanced companies in Asia.I understand that you use AI in a number of ways such as the response of the autonomous driving and that you also have AI in this part.Could you tell us which business you currently work with?Certainly.So we call AI the assistive intelligence.So we make sure that whenever human feel that part of our job is boring or is difficult to train newer generations of people because they find it too tedious to do,it's much easier to train machine assistants to serve as assistive intelligence.And this includes, for example, at the moment, I'm in the public sector.And because we're a democracy, we have to do a lot of public consultations.People write their ideas from the public commentary to all the regulations and all the like opening up the ocean, opening up the mountains.What about safety?What about indigenous culture?In every policymaking, there's like tens of thousands of people who want to share how they feel about any particular controversial issue.And previously, well, it's the frontline public servants understanding the qualitative input.But what if we can make sure that people can moderate each other's views and assistive intelligence can show people what are their common values,despite of their different positions, where we use the Polish technology that is exactly the application of AI designed for scalable listening for pro-social media.So people can think of social media not as something that divides people, but rather brings people together to do policymaking.It will have done many policymaking, including like our public diplomacy projects with the U.S.And regulation of Uber, of Airbnb and other platform economies.And also, as I mentioned, ocean and mountain policy and so on, all using this scalable listening AI technology.That is great.I'm very open.I'm very advanced, you know, our social technology.That is great.And so my next question is in Japan due to the impact of COVID-19.So effort to digitize the government and divide that has been emerged and added to the issue.So recently that was the news of the establishment of a new digital default.What do you think the key to making this platform will be?Yeah, the key is to think beyond IT, because if you think about IT or ICT,there's information and communication technology that's mostly about connecting machines to machines,which is important, but digital is about connecting people to people.And so if the digital agency can send itself mostly about machine to machine work,then it will not lead the direction for the emergent technologies.Rather, it will be playing catching up to the emergent technologies.However, if it focuses on the common values, on assisting the society,and getting the society's input on where the technology should go,as in Japan, we have heard that when industry is at 4.0,the society is already at 5.0, meaning that the society needs to lead the industry.So if the digitalization can be a breach this way instead of this way,then I think it will work much better.Your point of view is I think the right.So you said some people may worry that AI is going to take our job,but if AI is assisting intelligence, that's worry can disappear.That's a very good word I think.And so I trust that you will agree that AI needs to be the center of any data transformationand the technology driver that will assist the management innovation.This is something I learned while I was working as a tech giant at the event.What do you think about AI's role in helping management innovation?Definitely, I think when we think about management,we often think it as a way to manage uncertainties,to manage risk, to manage against unforeseen scenarios.But Japan, just like in Taiwan, we understand where earthquakes just happen.You cannot manage away earthquakes.You can't manage away typhoons either.What you can do instead is to be resilient,is to have advanced warning and when the disaster actually comes,not only to minimize damage,but also use that as the opportunity for co-creationand imagining a new better way forwardas a way to collaboratively learn togetherbecause that's where people focus their energyas you just mentioned because of COVID,everybody is now focusing their energy on digital transformation.So don't let a good crisis go waste is the main idea.And in this, if we have good assistive technologythat are value aligned,meaning that they agree with our valuesbecause we have the common values already somehow figured outduring the social consultation,then the AI can show us the likely futuresif we do the digital transformation this wayor that way and so on.And so they can assist us in seeing longer horizons.Also, assistive intelligence can help usto give accounts to each other.The social sector can be accountable to the public sectorsaying, hey, instead of just investing the moneypurely on reimbursement,we have a better plan as social innovationthat can make even more social impactand the same way, the business can also maketheir sustainability reports accountablesaying, hey, if you do business with me,it's not going to be only good to you as a customerbut also good for the societyand also for future generations.Again, AIs enable us to take those very tediousworks of accountabilityand turn it into something that just happensin the flow of workand it's also very good for managing the triplebutton line, not just the financialbut also the social environmental ones.I like Julian.That's very great.So next question isfrom your global perspective,what are some of the biggest challengesto digital transformation facing today'scompanies?And what do you thinkwill be the triggersfor digitalization and AIproject in this company?Yeah, I think I just mentioned the ideain the flow of work.I think the main challenge now is many organizationsincluding corporationsthink AI as something extrasomething externalsomething like a black boxsomething that you just investand it may or may not workand so that's not a strategy, right?it's more like a hypethat people understand about AIbut actually what's importantis a continuous investmentin both the leadership levelso that we understandwhat kind of work is bestfor robotic assistanceand what kind of work is bestfor human judgmentbut also to make sure that the infrastructureof the company or the organizationis built so that data isjust gathered continuouslyfor future transformation.many people who think in IT termsthink in the oldmetaphor ofextracting data, transforming dataloading datainto a database or things like thatbut nowadays with assistive intelligenceit's the reverseyou first just load the dataand then you think about how to transform itto make the insightsit's just likein Cyclopedia Britannicayou write the articleyou edit the article and then you publishthe Encyclopediabut in Wikipedia you first publishand then people who don't like your biaswill just contributeand to make their way into the better qualityof the report.so the same goes for the co-creationof the data in the assistive intelligence era.Thank you very much.and the next question is thereif you were the best ownerwhere would youstart with AImanagement and could you give usexample 2?Definitely.I think what's most importantis making sure that there'sa digital twinto all theincluding the human resourcewhich I don't really like that wordbut anyway HRand then auditingand then also salesand many other departmentsbecause if you do not have a digital twinyou're essentially working with two setsof data, one that's recordedand one that's testedand the tested knowledge of course cannotbe fully expressed or capturedbut if you build a digital twinso that in the flow of workfor example when we're having this conversationwe both hit recordingand that created a digital twinof our conversation that could bereplayed and turned into AI basedcaptions, analyzed, summarized and so on.but if we meet face to faceit takes an extra effort for usto think about the recording possibilitiesbecause it's an extra deviceI have to actually use thisvery bulky deviceto capturethe face to face meetingand so in a sense a teleworkingculture makes this ideaof capturing a digital twinmuch easier than face to faceworking environmentsso I would systematically look intoas I mentioned all the HRaudits, sales, M&Awhatever departments and seewhich part of it are not capturedby this digital twin thinkingand maybe use teleworkingor even virtual and augmentedreality to make sure that peopleare comfortable working with eachother's digital avatarsand so yeahif that case in terms ofhuman resource developmenthow do you think business executiveand managers shouldmultiply approachmanagement on the AII think one of the main thingthat I worked withthe government leadership hereis making surethat they useAI in the flow of workat least a little bitso that they understand what they liketo listen at scaleas I mentioned using the polis technologyor to have realtime captionsas we are having a conversationor to use AIto detect anydifficulties in like solvingwater pipe leaks and so onwhich is a real problem this yearbecause we don't have a typhoon this yearand so the rainfall is somewhatshort and so we use AIto diagnosethe SCADA data of the operationtechnology data to find outwhere to repair firstwhere it's most likely that has thedaily leaks and so onbut it doesn't have to be deep learningit could just be AIof the 20th centurywhich is the olderlike linear regressionand whateverclustering, k-means clustering,principal component analysisthey are not deep but at leastthey are very easy to make account ofit's easy to explain how they worksas compared to deep learningand it's easier to showthat hey it's really solving a realimportant problem in realtimeand so once they get into thishabit of working with realtimedata and as ifintelligence then if we do needdeep learning or other technologythat we need more investment inmaking sure that it's accountablethen they will be comfortable making that decisionotherwise just jumping intodeep learning, I don't think that worksSo, as youproposed pastpair from wherethat's a really great wordand soI'm sure that thisconcept can be applied to the futureof such celebrity in Japanand Japanese companiescould you tell us your thoughtsabout the past which isespecially important for businessYes, soin many casesthe business hasan emergent situation where it needsto respond but it takesa world of differencebetween responding in 24 hourstime or respondingin 24 secondsif you can respond to a new situationin 24 secondsit means that the analysis worksin near realtimeand when the analytics works in nearrealtime then by the timethat you are aware of the situationyou're already thinking about resilienceyou're already thinking aboutthe whole picturebut focus on this firefighting modewhich will be what happensif you only get notice about it24 hours laterso, this near realtimeresponses them in Taiwanfor example in our central epidemic we rely on this very old technologycalled a toll-free numberthat anyone can call 1922and say that an emerging situationhas happenedrelated to COVIDof course they can also use chatbotsand other technologies toll-free number makes sure that everybodyhave equal inclusive accessso, we invested a lot of budgetand time to make sure that thiscall center can escalateany new situation to thecentral epidemic command centerwithin a couple minutes at mostand so this, I think, is a very good wayto get people into the habitof working with the whole pictureinstead of just on firefighting modewhich would be what happensif you don't work fastthank you very much for your messageand my last questionis your final ideaso, Peter Japanwill work with Japanpassionate andinnovate businessreaders to create a new futurefor Japan in AI and digitalmay ask you providesome message for Japanthis business communitysure, definitelyso, nowadayswhen you talk about AIthere's no escapingof the question of data ethicsof whether it's ethicalto over-concentrate dataon the statewhich will enable state surveillanceand authoritarian intelligencenot assistivebut or to concentrateto some very large companiesmultinational companieswhich will enable surveillance capitalismand anti-competitive behaviorand so if you think about this askind of a pitfall on the leftor at the pitfall on the rightwhat's important is that we walk the middle waybetweenthe state surveillance on one sideand capitalist surveillance on the otherthere is the middle wayof co-creation of assistive intelligencewith all your stakeholdersincluding everyonein your supply chainalso all of your customersand also all the potential stakeholdersin the business that you doif we work on this joint datacontrollershipso that everybody feels they have a fair shareof sayin how the data is governedand usednot only can we pave a third middle waybut instead of left wingand right wing we can have a up wingwhich raises the whole societyupward instead of just a few peoplein the governments or in the business worldand so that was my last questionthank you very much fortoday's sessionand so thank youthank you for the great questionslive long and prosperthank youAIをアシスティブインテリジェンスとトラエナオス