現在 女士們和你們的擔心開始的活動開始了現在 女士們和女士們接下來是第一集的社會新興建築物主要的主持人是美絲奧德里頓新興建築物的副總裁所以 女士們和女士們你們準備好了嗎第一集今天 準備好了嗎對 好的我知道我們會有一個很棒的旅程你相信我嗎是的開始的旅程女士們 請給我們一句鼓勵歡迎美絲奧德里頓新興建築物的副總裁大家好很高興你來這裡我相信預計會在一分鐘內我們開始預計開始了讓我先記起謝謝所有的副總裁在這個活動中我們看到很多的能量包括 展示我們可以聽到的能量我會請你參與我所在的討論我會給的所以 最簡單的想法就是 如果有電話你可以聯絡到這個網站 叫Slydo叫Slydo.com它應該會在一分鐘內或者是的所以 你可以 either scan the QR code或者 你可以去Slydo.com然後 進入現場前面0 你不必要進入帳號所以 01115如果你在Slydo.com或者 你可以 scan the QR code是一樣的然後 你在這個帳號你是一部分的聯繫你可以問我任何問題在下一 40 分鐘內你也可以喜歡每個人的問題問題是 最多的喜歡會掉到上面然後 新的問題會在下方這個方式如果你看到一些問題你也會喜歡你不必要再回答最好的問題是它們都是獨有的所以 誰知道誰問了問題所以 我們可以問很多很直接的問題每個人都會用電話所以 我會回答這些問題所以 第一個問題是我的日子我的日子開始了我很高興 to be here and speak with you我會開始 with a short presentation然後 我會回到所有的問題你們都會聽到關於 digital social innovation所以 開始 with我會想和你們分享我的辦公室這是我的辦公室在台灣的國際政府辦公室這是叫做 social innovation lab是在建國花市所以 每個星期建國花市和 Jade Market opens所以 如果你們想在週末這是一個很棒的地方然後 旁邊的建國花市是這個地方是 co-created by hundreds of社交新興建在台灣我們在這週末就在我們的第一年祝福了所以 你可以看到這裡的標誌是很有趣的因為 人們是瘋狂的看起來 人們是瘋狂的看著世界 從不同的標誌來看然後 他們說了這個標誌的語言更好 更好所以 這是為何我們在社交辦公室做了他們的獨有的目標讓世界和整個世界加入他們的目標進入這個空間的設計所以 這點是 co-design和 co-creation我自己學會了我15歲的時候那時候是1996年然後 社交辦公室就開始了我告訴了教授還有主持人說這個社交辦公室就開始了我可以無論是在讀書或是在10年內發表資訊或是我可以參加這個 AI 研究然後 創造10年後的 textbook在未來的所以 我想要離開 junior high school然後 開始我的最初的 startup enterprise然後 驚訝的所有教授都 agree with it他們也想要參加所以 這點是台灣的新的教授當中教授和教授同學同學就離開第二年 junior high開始 startup together然後 我參加了這個社交辦公室和網絡辦公室然後 都參加了這個神經病的 idea of radical transparencyradical trustand voluntary association這些是教授我學習的當我開始我的第一 startup在社交辦公室在最早的90s所以 我把這些精神為台灣的第一次教授而在社交辦公室改變了所以 當我們說到 co-creation我們並沒有代表這個社交辦公室的是 co-created by innovators就像他們說我們要有一個 chef所以 我們要有一個 chef它開放到11pm every night所以 當我們吃飯人們可以連接在一起只要你做任何的 sustainable development goals here你可以用這次的教授是 co-working space是 event space然後 他們問我是我的每星期都在學習所以 我每星期都在10am-10pm是我的辦公室所以 任何人在學習可以跟我說我可以跟他們說他們的 whole transcript都在網上發佈了給大家使用所以 每個星期這個 space可以改變然後 幾個月前我們有些特別的朋友從 MIT Media Lab這些叫Persuasive Electric Vehiclesor PEVs而這些 PEVs是自行駕駛車他們很小他們有這個高度他們很快就不會動所以 即使他們走進屋子裡沒有 harm done最好是他們的 open source和 open hardware所以 每個學生和每個 co-creators都可以看看看能否幫助例如 藍燈在建國花市去 pick up flowerspick up pots of flowers然後 把這些 companion animals然後 他們會追你走然後 最後你能直接跳上去然後 駕駛回家等等所以 這樣我們學會了 AI和 集中性能力當我們看 AI我們不想想到有什麼在空中的領導之類的東西我們想到什麼在我們的每天體驗中如果你們不喜歡這個程度他們不知道如何去行動這個情況會發出藍燈如果你們不喜歡這個程度你們可以跟教育學生跟 C.S.E.E.E.的能量然後 變成一隻狗一隻狗等等所以 這樣我們建立新的社會法律如何進行 AI進行社會這就是一種建立的精神而 為什麼我們不想這件事變成例如 建立一個經濟以及一個商務在其中或是民生技能在其中還有一個環境保護在其中因為這是天氣不好的近年更正的貴重性我們有人們派對一 二博士還有其中一個精神圖中間的都是大部分首研觀念我們沒有看到他們但是他們在壓力來做造成akin的人觀念各大各大各的東西各地的公司和公司各地的東西但是在這段時間尤其是在社會網站的新的社交網站這沒有任何工作這是壞的模式首先因為你不需要一個公司或一個市長去辦公室任何人在對的 hashtag可以辦公有幾百萬人沒有必要的再活動公司第二件事是有很多的東西不是只用機器學習但也有分辨的文件還有每個行程我們不能有一個公司或一個公司每個電腦技術都不合理所以我們改變了我們叫做不同的公司模式我們叫做合作公司不僅要問如何去承認人們如何去承認公司我們現在問大家有不同的位置那是可以的但我們可以找一些普通的資料嗎然後我們找出這些普通的資料讓大家可以說這是好價值然後我們問政府做了什麼阻礙這個價值從完全實現自己它可能會出現法律或法律或政策它可能會出現一些資料的方式但我們問根據普通的資料能否有人做一些新的改善讓很多人會解決這個問題給所有人然後離開其他人所以這種構思叫做社會新的所以要舉個 concrete example如果大家在台灣如果你們在開始想想他們會做這個社會更好但我們的 existing規律或法律總之會保護那些事發生你們可以去這一個網頁sandbox.org.tw找出一個公司一個公司就像這樣我想嘗試我的想法但規律或法律現在正在寫成這樣阻礙我從實現社會的能力所以我想改善給其他人所以你要提供其他東西或者是罰款或者是法律的 existing regulation那就是它們的 take每一件事都是自動兼職免費的一群譚德的職業律師會作出哪個權力是負責的如果這個社會經濟 國際社會國際社會會議會如果這是A.I. BankingFintech Innovation國際社會議會如果這是關於自動車那是更快的比自動車更快的國際社會議會我們預期這條法將會通過在1月和1月我認為U.V.會通過如果需要5G broadband我們會通過這條法因為國際社會議會我們會通過在1月和2月和1月我們會通過在1月和2月我們會通過在1月和1月我們要求你解決法律的規律讓社會有更好的想法更是我們的規律和規律包括商業模式它可以在一年後嘗試如果社會認為它是一個好的想法它可以在一年後減少它可以在一年後減少它可以在一年後嘗試在二年後減少如果社會仍然認為它是一個好的想法我們會被迫使用如果是一個規律去修改你的版本兩個月後兩個月後修改你的版本去修改你的版本去修改我們的規律所以你會變成基本上是一個規律的副創作如果這需要法律改善我們的議員會需要更多時間所以三年或四年內他們會被修改但是在那一段時間你基本上會有一個模式因為你的體驗可以繼續走包括商業模式至於當地的政府將會在最好的決定之後參加者會進入市場但是如果沒有即使社會不認為它是一個好的想法還是 大家都會學習到什麼我們要感謝那些投資者在描述時間也要提供資訊讓那些新興建築者可以嘗試從不同的角度現在的 logical question is我們如何知道人們需要的東西我們如何知道人們在不同的地方他們關心的事情這就是為什麼我每天都在台灣每個星期都在台灣每個星期都在台北每個星期都在I mean the indigenous placesrural placesremote islandsthe places that really needthe technologists like youto introduce your startup ideasto work with them to improve the societySo for example this is Huadianand more remote places likeTai Dong can also teleconference inand I go thereI live for a couple daysI work with the local innovatorsbut the most important thingis that all the 12 ministriesare also still in the social innovation labenjoying the lovely geometryenjoying the excellent foodand a very relaxed atmospherebut all the 12 ministries are thereand see through my eyesthrough two-way live streamingwhat the local people's faces arewhat their stories areand when they raise a questioninstead of like in the bad old daysthe ministry of interior will sayoh I have to ask the minister of financethe finance saysoh I have to check withthe minister of health and welfareand five months will have passedbefore you get anything from thembut this timeit's impossible to do things like thatbecause everybody is literally in the same roomso there's no way for peopleto shuffle things aroundthey can just brainstormand get back to you at mosttwo weeks after each regional meetingafter each regional meetingwe publish everythingto the public internettwo weeks afterwardsso every time people can continue the conversationand through this waywe co-created hundreds of solutionsfor longstanding lasting problemsbut if after two weeksall the ministries admitthat the government at the momentdoesn't have a good solutionto this social or environmental problemthen you can use this as the costas the rationaleto file your sandbox applicationbecause the government has essentially admittedthat we cannot fix thisat leastuntil the next budget yearso this is a roomfor you to step inand to be involvedand so for exampleif this is aboutself-driving vehicleswe have a public exhibition spacethis is just rightoutside the high-speed rail station in tainaneverybody can see a simulationof a self-driving vehicleinteracting with motorcycleswith a lot of differenttraffic scenariosso everybody can feel safeand co-create the social normsand we make that social norminto the technologyso that they're aware by designand safe by designand so at the end of theexperimentation periodwe'll have to make a judgment callof whether the societyfeels safe about itand how do we do thatif you have a sandbox experimentthat runs for a yearit may involve thousands of stakeholdershow do we actually listento thousands of peoplewell we use AIto listen to peopleso this is AI power conversationcalledpolisand this is open source technologywe've been using it for three years nowand the conversation you're seeing hereis the first usewhen we moderatedhow uberx enters Taiwanand so a conversation like thishas four stepsthe first oneafter a year of experimentor crowdsourced open datawe make sure everybodyknow what the facts areand when we say open datawe don't just mean open government databut citizen scientists as welland after we can agreeon the same set of factswe just have a monthto check people's feelingsfor the same factI can feel happyyou can feel doubtfuland it's all okayand the AI power conversationis just for feelingsbut after people findthe common feelings and valuesthen we can brainstorm on the ideasthen best ideasare the one that addressthe most people's feelingsand then if the ideaswork together coherentlywe turn them into a judgmentthat is turned into the lawso for examplewhen we consultedfor the autonomousfecos sandboxthis is how it looks likeeverybody gets a linkon your phoneand you see your avatarherethat's in the blue dotsand you see one fellow sentimentfrom one fellow citizenand you can click agreeor disagreeand as you click agreeor disagreeyour position will moveamong the Facebookand Twitter friendsthat you haveor if you choose not to sign inthen among famousFacebook and Twitter friendspeopleand that's allso thenafter you click agreeor disagreethe next sentimentwill appearand after doing sofor a whileyou can also shareyour own sentimentthe feelingsfor other peopleto resonate withand this has two effectsfirstbecause you can seepeople on the other sidethey're not your enemiesthey're not polarizedlike ideological enemiesthey're your friendsand familyyou just didn't talkabout this over dinnerbut now you knowtheir positionbut the second thingis thatpeople's positioncan changebecause there isno reply buttonall the spacesthat we builtit could beslidoit could bethe e-participation platformeverythingit could bepoliswe don't havethe reply buttonand the reasonwhy isbecause whenwe seereply buttonwe seepeople attackingeachotherwe seepeople postingcat pictureswe seepeople derailingthe conversationbut ifyou don'thave thereply buttonif youdon't agreewith thethings on itthe best thingyou can dois justto proposesomething betterfor otherpeople tovote onso itcan onlyadd toeachotherbut youcannottake awayfrom each otherand soeverytimeafterwe run thisconversationwe seeand thisisvery counterintuitivebecauseif youyou justlookatmainstreammediaorevensomesocialmediayouwillthinkthedivisivestatementsareallthereisbecauseitdominatestheentirediscussionbutusingAIfacilitatedtheconversationwe findpeoplespendfarmoreenergyoncreatingthestatementsthatareofconsensusthateverybodycanlivewithandworkwithandsowebindourselfstakeholderconsultationafterwardsandsothroughthiswayweintroducetheideaofsocialinnovationbygettingpeople'sfeelingstogetherchatthecommonfeelingsandinnovatebaseonthefeelingsandsofinallyifwecreatesomethingthatisjustgoodenoughandasocietythinkthatitcandobetterwewe also work with a peculiar way called GovZero. I'm part of the GovZero movement. The movement started around the 2012 so it's been around for six years now. And the idea is very simple. I encourage you to bring this idea back to your country. We just had GovZero Italy last week. So the idea is very simple. Say you have a government service or website. Like the legislation is LIGOVTW or the national budget.Or whatever, right? The environmental agency. I'm sure it's the same in your country. Here in Taiwan, all the government services and websites ends with GOV.TW. So very simply, GovZero is just a domain name called GZeroV.TW. And so whenever anyone in the civil society see a government service or website that they don't like, they think they can do better. They can build a shadow government website with exactly the same address. You just change the O to a zero in your browser.You don't have to Google for it. You don't have to send out Facebook advertisements. All you have to do is to change an O to a zero. And then you get into the shadow government. And the shadow government is more interactive. It's more useful. And the good best thing is that in the GovZero movement, we all relinquish most of our copyright.So by the next procurement cycle, if the government thinks it's a good idea, if the people likes the idea, it just becomes the official government website. For example, the inaugural budget.GZeroV.TW is a visualization of the national budget. You can drill down to each and every budget that you care about and have a real time conversation around that budget item.Well, now we have merged into the national e-participation platform. Join the Gov.TW. And so on join platform, you can now see all the 1300 ministerial projects, their KPIs, spending, procurement, progress, everything. And in each and every budget item, you can type in any question and a career public servant will just answer you publicly.You don't have to go through MPs. You don't have to go through ministers. Everybody can have a fact-based conversation around the budget item. And that was the first GovZero project that's merged now into the e-participation platform.And the platform has now 5 million active users out of 23 million people in Taiwan. So one quarter of the population is pretty good. And so we're spreading this idea as I mentioned to Italy.So now if you go to budget.GZeroV.Italy, you also see the Italy GovZero because we don't have a trademark or patent on this GovZero idea. You can try it out in your own country as well.So before getting back to the questions, I want just to show you with one GovZero project and then a poem. And then I'll go to the questions.So this GovZero project is called Airbox. Some of you may already have heard of it. It's one of the ideas that's really caught on in Taiwan. More than 2,000 people installed this very cheap, like less than 100 US dollars.Air qualities measurement devices on the balcony, on their schools, on their homes and so on. So they get a measurement that's closer to their everyday living.And then it's not just for personal use because there is an academia-cynica project that is a network that everybody can upload to the cloud and the cloud displays.So you can see at a glance the digital divide in Taiwan. Well, no. Well, you can see at a glance the air quality in Taiwan.And so you also see that people really donate and contribute those sensors, those air quality measurements without any intervention from the government.And so in other Asian countries when I share it to other UN delegates, they all tell me that in their area of the world it may not be possible to grow to 2,000 people.If it grows to 100 people, the leader will be poached into the government. If it grows to 500 people and the leader refuse to join the government, then maybe they will get disappeared because it really challenges the legitimacy of the central authority.If you have two numbers published by the number of the environmental agency of the government on the first hand and the one that you measured with your friends on the second hand, even though this is more precise,of course you're going to trust the one that you and your friends measured. So this really creates a legitimacy challenge to the central government.But in Taiwan we're very unique. Our central government says we can't beat them, so we just joined them. And so we embrace officially all these input from the citizen scientists into our national collective intelligence platform.And we work with them saying, okay, so there is some digital divide. There are some places in the mountains and the more indigenous places that people are not contributing the citizen science measurement.So we will do that. And we also hear from our citizen scientists that they really want a measurement device here, right in the middle of the Taiwan Strait, because they want to tell the domestic versus non-domestic air flows.But they cannot really put an air box there. Even with drones, the drone can run out of battery. But we can because we're building wind-based power plants on that place.So we're placing their technology on the state-sponsored wind power plant. And so through this way, because it's all open innovation, it's just on GitHub.You don't have to ask for a license. You can just download, put on an Arduino, put on a Raspberry Pi, and you too can do this in your country.So you can also host a data analysis yourself. But if you don't change the default website address, it just goes to Taiwan.And so we have an international network that just reports back the numbers to Taiwan.So if you're interested in this cross-sectoral collaboration, you can go to collectiveintelligence.taiwan.gov.tw.And there's many people who tells us, well, how come the environmental protesters are willing to put their numbers to the National Superhype Computing Center for AI analysis?Are they not afraid that the government will change their number?And we're like, no, because we work with people in the distributed ledger space.So before they upload, they take a snapshot and store it to the IOTA distributed ledger.So it's like a blockchain. And once it's on the blockchain, if we try to change the number, everybody will discover that we changed the number, so we will not.So this is a very simple emerging technology that makes sure that everybody is honest, that has reliable data.And so through this way, with the digital technology, not as working on just one or two of the sustainability goals,we don't see the economy, the environment, and society as something that is in tension with each other,but actually with availability of reliable data, especially with distributed ledger,with effective partnership, especially between the collective intelligence and artificial intelligence,and with open innovation, especially AI that feels personal, like personal computer,that everybody can learn and modify,we truly believe that we can create a world that is free of the old bad days,where the silos, the different sectors fight against each other,but rather we can create something that is of common value to each other.And so finally, two years ago when I entered the cabinet as the digital minister,people asked me to write a job description,so I'm going to read the job description to youand then go to your question for the next 20 minutes,it goes like this.When we see the internet of things,let's make it an internet of beings.When we see virtual reality,let's make it a shared reality.When we see machine learning,let's make it collaborative learning.When we see user experience,let's make it about human experience.And whenever we hear that a singularity is near,there's always keeping in mind and always remember that the plurality is here.Thank you for listening.So we have 22 questions and about 20 minutes,so I'll spend maybe one minute for each question.So what are the sustainable goals and how the AIS can help the sustainable goals?This is a great, great question.So the sustainable goals,which you can read up on globalghost.org,is a set of 17 goals and 169 targets that the United Nations signed into effect in 2015,that's three years ago.So the idea is this very simply for three years or so,the UNDP asked the entire world more than one million people.What is the kind of the world that you would like to see by the year 2030?And they get more than one million voices.And using,I'm sure,some automated augmented intelligence,but mostly diligent UN helpers,they get all those million voicesinto the shape that reinforce each other,rather they take anything away from each other.So you can see if you care about,for example,maybe care about industry,innovation and infrastructure.And it calls for the universal access to internet.It calls for domestic development and industrial diversification.It calls for enhancing research,enhancing access,especially mismiesto financial services and things like that.But they make sure that each and every of those targets,they're all supporting each other.They're not taking away from each other.So these are kind of like the common values of humanitythat everybody has agreed to by the year 2030,we have to get there.So those are the global goals.Of course,it doesn't say how to get thereand that is where AI can helpbecause a lot of the sustainable goalsis about making innovations instead of trade-offs.Previously,if we think in a zero-sum way,if you want to save somebody two hours of time,like in public service,somebody somewhere else in the public servicewill have to spend two hours or moreto save you two hours of time.It's a zero-sum game.But now with AI,we can look at any part of our workthat is trivial.Trivial is a computer science term.It means that no matter who does this task,it only differs in accuracy and speed.That's all.So this is fixed input,fixed output,and only differentiation is on the precision and speed.So it costs for no communication.It costs for no life experience.It costs for no empathy.So in all our daily work,there are some moments like thisthat we may not be aware of it.But now if you work with experienced designers,if you work with professional consultantsor if you train yourselfto be more aware of opportunities,then everybody can identifya few hours in everyday's lifethat are trivial.And these are the onesthat we can automate with AIs.And which is why I always thinkof AI as augmented intelligence,not artificial intelligence,because it's really just augmentingour daily lifeand automating away the trivial parts.And that enables SDGsbecause then we can make everybodysave a lot of timewithout creating externalitieselsewhere in the world.And that is the spirit of the common goals.And I can go on and on,but it is a basic idea of it.Two people would like to knowco-creation is essential to drive innovation.That means making tech open source.But doesn't this meanoriginal ideas lose competitive advantage?This is a great question.This was a big debatearound maybe twenty years agowhen a term open source was firstpronounced.And I was part of that movement.But now I don't think anybodydispew this anymore.Microsoft has donatedall their Linux-relatedpatterns into the Open Innovation Network.Microsoft is now GitHub,which is one of the largestplace where co-creation happens.IBM is now also Red Hat.And so what we are seeing nowin the across the industryis that people are seeingthe algorithms, the codeas not exactly an asset.It is more like a sharedinfrastructure, a shared fabricthat everybody can maintain together.And so everybody don't have tospent a lot of energyon keeping it up to date.People are actually treasuringmore about the relationshipsthat they build through.For example, data exchangewith the people that they care aboutthe content of their product.But the code itself is lessand less of an assetand more and more of a burden.Which is why more and more peopleare embracing Open Sourceas a way to share the maintenance burdenof innovations while keepingits competitive advantageof tight relationshipwith the people in the stakeholder baseof that particular enterprise.So this is what we are seeingin terms of Open Innovation,especially around software.This is still under debate.In the science publishing,Open Access is stillmaybe what we were ten years agoas Open Source movement.But gradually we're changing the waythat people access and shareand create knowledge.Three people saidlove this GovZero idea.How do you ensure it doesn'tpropagate a disinformationphenomena like the one thatplayed Facebook during the last5月?So as I said,if you don't have a replybutton,there really is no wayfor other people to attackeachother.And this is themost important part of it.And there are also existingGovZero project that workswith disinformation.I'm going to just show you onevery quickly,maybe a coupleminutes.This is calledco-facts.For the onewho don't readand the name ofthe project in Mandarinis literallyis it true or not.So what's it?It'sbasically a bot.And it starts fromline.Line is a popularit's like whatsapp,N2N encrypted channel.It's not gender biased.Everytime you refresh,yousee a different person.Once you add it as your friendeverytime you see a rumoryou can forward to the bot.And the bot does a factchecking for you and get backto you whether this is trueor not.And the crowdsourcefactchecking is entirelytransparent,it's in the open.And so what we see hereis all the trending rumorsin the dark encryptedend-to-end encryptionchannels.And I will not keep you ontoo long because it kind of pollutesthe mind,but I only want youto know that all these getsfactchecked within minutessometimes,but almostalways within a few hours.And so what this does isessentially it surfacesevery singlerumour that is trendinginto a publicurl so that you can bothsee the clarification veryquickly,but it also itselfis a social object thatpeople can have a realconversation before it isjust a rumor on theend-to-end encryptedchannel,but now it becomesa public url that everybodycan have a conversation.Onceit becomes that it loses itsvirality,it's likeinoculation,it's likevaccine.Vaccine isessentially virus thatare injected to peoplethe civil society helpseverybody to fact checkwhat is currently trendingthat this information.So thisis also innovation that Iencourage you to bring backto your country,far as I knowpeople are working on whatsapp andother versions of the samegov-zero idea.Do Iconsider myself a rare breedin a town government?Am Igetting enough support from thegovernment?Yes,of course.SoI joined two years ago andmy staff is very uniquebecause my staff isliterally I can poach oneperson from every ministry inthe cabinet.There istwenty-two peoplein my office now.There'sthirty-four ministriesin the Taiwanese cabinet.Somaximum I can havethirty-four coworkers,butnow I have twenty-two.There'ssome ministry missing,like Ican have some ministry ofdefense for some reason.Butin any case,maybe becauseI don't touch state secret.Butin any case,thiscross silo work style isvery important for theministries because then noministry dominates thediscussion of the digitalpolicy,but everybody canpropose freely.And so weourself use digital tools.Soif you go to ey.pdisk.twand platform,that any appthat runs on it iswe had six monthsof penetration testing bythe top-notch white-headhackers that are second placein DEF CON worldwide and so on.Sowe know that this is safe.Everyday I wake up,I look atmy twenty-two colleagues fromdifferent ministries and theyall work out loud,meaningthat they're not afraid of lettingother ministry know whatis going to happen.Sowe track everybody's workvery easily and form ad-hocteams.Now for you whowork in a startup world,thisis nothing,this is justajal,kamban,skrum,whatever.Butthis is career publicservice we're talking about.Thisis the entire centraladministration,all atleast twenty-two ministrieslearning to work in averyhorizontal way.Andhorsizontal protest,horsizontalorganization,they're notafraid of it because theythemselfhas been part of ahorizontal organization.Isthere any work onlevel three to level fourautopilots?Yes,there'smanyreal needs fromthe municipality governmentthat they tell us.Forexample,the Gaussian citygovernment really want to takethe last mile of the metroit's not yet built.Ormaybe even in the love riverthey want to have someself-piloting boat that alsocounts as level four,even thoughit's not a car,it's a boat.Andin many places we're alsohearing needs forautomated drone delivery.Andall of this will beenabled by the self-drivingvehicle sandbox.Rememberthis is not a minister oftransportation,this isa hybrid vehicle.Itcan be,I don't know,hyperloopor something.Anythingthat people's imagination takeyou forward,you can apply forthe sandbox and it'sexpected to be passed reallysoon like this December.Howdo I think about the AIenvironment opportunitiesin Taiwan?I think Taiwanin terms of research hasalways been one of the bestplaces to produce qualityAIresearcher.Well,stillam starting 1996or something.And a lotof my friends around thetime already work onmachine learning.Even deeplearning was not in thetrending situation backthen.But what I observedis that over time many ofmy friends just went tothe Silicon Valley or toLondon or to some otherplaces to continue theirresearch.But in the pastthere were a lot ofmachines in Taiwan.And thereason why is that whilethe previous generation ofmachine learning mostlyfocusing on abstractactivities like playinggo or playing chess,theseare abstract,but nowadayswe're seeing AI'sembodiment in a sense thatthey're now finding abody in the robots or injust the air qualityof the world.And for thispure software doesn'twork anymore.You have tohave a very good designecosystem.You have to havea very good supply chainICT ecosystem.You have tohave good optics,goodaudio arrays and thingslike that.You also haveto have a very quickturnaround.Even what wecall AI on chip must bevery quickly prototype.Andyou can build a fullyembodied AI without muchhelp from any overseasprocurement sources,whichis why,like Microsoftsetup AI lab,I think 100up to 200 people.There'sa director of AIspeech research,KortanaEason Du,who went backto Taiwan from Microsoftis now building more than100 people working also onsmart city and humanstate.Literally from everybodyIBM,NVDA,Google,Uber,you nameit.There's all AIexchange programs and ourresearch labs now in Taiwan.SoI think this new trend of AIembodiment is really bringing alot of talents back toTaiwan.And also we're veryeager to try new ideas.Sowhenever you have a newidea that breaks the lawor regulation,you can justtalk to me every Wednesdayface-to-face in the socialdata and algorithmbased law can work togetherin a way that is fairto both sides.Sosomeone asked,isAI best if it resembleshuman,or is AIbest if it resemblesmachines?This is agreat question.I thinkit's somewhere inbetween.So humansactually a lot of humangestures that we thinkare human.Likewhen we smile,we followeachother's smile.If Ilook somewhere,you followmy gaze to somewhere,andthings like that.These werenot early hominid trades.Thatis to say we evolved thesetrades with dogs.Theearly wolves and the early hominidscoevolved these traitsfollowing eachother's eyes,followingeachother's nonverbalexpressions,which is whyeveryone even people whodon't know dogs at all,whenthey first time hear a dogcry,they can get emotionin the dog's communicationbecause we have coevolutioncircuits in ourneural network,then the samefor the dogs also.Sowhat I think is mostimportant is that we thinkAI as augmented intelligenceas something that is woveninto the social fabric.Andif it's not humanlike,we don't expect adog to suddenly become afull-fledged human,but weexpect it to adopt especiallynonverbal human traitsso we can understand their emotionbetter.So we can,forexample,not only theycan know more about usin our society,but also wecan know more about individualAI agents and the situationtherein.And there's a lot ofcapacity for this kind of empathythat's built into our brainthat I think the future humancomputer interaction canreally fully use.SoI think this I already answeredso I will just talk aboutweakness.We have a lotof strengths,but our weaknessis at the moment of imaginationand risk taking.Becausepeople in the earlier generationspecializing in hardware often think of startup assomething that is very costly,veryintensive.So whenteenagers now want tostart new startups,sometimestheir parents say maybenot go there,maybe afteryou become financiallyindependent.But now if youhave a good idea in AIactually you can crowdfund,youcan crowdsource most ofthe supply chain and most ofthe work.It doesn'tactually cost much tofail.So it's entirelytheir direction we're takingand the newer generation aretotally for it.So whatwe're now downdoing is doingupward management of ourupward generations to convincethem that startup doesn'tmean the same thing as whatit meant in their generation.Astartup here is now justtesting an idea and ournew company act also fullyreflects this understanding.AndI must say sorry for themuch.Okay.Thank youvery much.And nowhere we are going togive you,wait a minute.Weare going to give you veryspecial souvenirs.It'sjust as you can see that.Yeahcause everyone is going togive you the tailormade pen.I'm going tointroduce it a little bit.Yes.It'svery special pen becauseit was shown in thenewspaper and magazinesand it's a tailormadefrom the designer ofTaiwan.And it has a prettyname,the weight of thewords.You know,allAI,we don't think aboutAI,it's all about thecomputer,right?And wetaught in the computernowadays.ButAI is good,but wehave to remember theworm that will reallywrite something with ourownhand.So,Evan,let'ssee the camera here.Thankyou for your sharing.Thankyou.And thank you,Evan.Thankyou very much.