IGF 2017 - Day 2 - Room XXV - OF42 AI Solutions for Nimble Public Services

 

The following are the outputs of the real-time captioning taken during the Twelfth Annual Meeting of the Internet Governance Forum (IGF) in Geneva, Switzerland, from 17 to 21 December 2017. Although it is largely accurate, in some cases it may be incomplete or inaccurate due to inaudible passages or transcription errors. It is posted as an aid to understanding the proceedings at the event, but should not be treated as an authoritative record. 

***

 

>> MODERATOR: Good morning, everyone.  I think we can start.  Welcome to the session, AI Solutions for Nimble Public Services.  

We have inspiring speakers and active participants.  So while we get to it here, it was noted by Peter from the foundation sometime ago.  He said that AI is probably the most important technology we have created.  It promises to bring changes and in every aspect of our life and government is not an exception.  So nowadays in time of fast changing technologies, user habits and official and public service in constant pressure and to perform better and linear approached to e‑Government is not enough anymore.  So we need some next level approaches.

And cognitive technologies, AI, offer some next levels too, which we can apply in different aspect in how the public sector works.  For example, as chat box, multilingual or road safety, you mentioned that AI can provide tools in different aspects in the Internet government in different aspects of how government performs.

And government also owns a huge amount of data that can be used for that.  So let's ‑‑ that's about our final here today, to look at the potential of AI in government sector and how it's already applied in government sector in different aspects.

So how we will approach this subject today, important aspects that was mentioned and stressed yesterday in opening keynotes was partnership.  Partnership is essential.  Industry government, academia and other work together to fully uncover the potential.  So for that reason, we also have a truly multi‑stakeholder panel here today during this panel today, we will cover such aspects as AI and public service delivery from government perspective.  Then it will be followed by insight to multi‑stakeholder framework of cooperation, how to connect the dots, an example approach, how academia, industry and government can be made and created and then it will be followed by AI application in public service and delivery in government.

So I'm pleased to introduce the, Dr. Signe Balina, President of the Latvian Association.  And Mr. Edmunds Belskis, Ministry Responsible for Digital Government, and state CIO in Latvia, and Dr. Andrejs Vasiljevs, and Mr. Aigars Jaundalderis.  We will have four interventions which will be followed by Q&A session.  So questions and discussions are very welcomed, and also online participants, we encourage you to use opportunity and pose your questions in WebEx, and your voice will be catched by my colleague Elza, and we will discuss these questions in our Q&A session.

So we are looking forward for interesting panel, presentation discussions.  Let's begin and let's start with Edmunds Belskis.

>> EDMUNDS BELSKIS: Thank you very much, good morning, ladies and gentlemen.  ICT professionals, governmental and public sector representatives.  As was mentioned in the beginning, it is important when we use artificial intelligence, not only in the private sector where you can see a lot of governments, especially E‑governance must use these technologies to serve more better services for our citizens.  And citizens as well citizens with a special needs are specific area to focus on, to make this life unique for all parts of society.

So to establish and improve artificial intelligent technologies, it's important to have also some preconditions.  If we speak about Latvia, we have excellent preconditions to move forward after utilizing paperwork which is done in several countries.  So the infrastructure is very important and Internet access, WiFi access everywhere, as well habits of citizens using new technologies, for example, in USDA countries, the latest surveys show that our country takes second place using broad and mobile datas.

So the next important aspect is how the society is ready for new technologies, and how population inters with the government, not only we see ‑‑ not only in the private sector, we have approximately 98% bank transfers made by Internet banking technologies, but also how the citizens use public services and how they operate with the eservices online.

As well legislative framework, which obliged governmental and the public service and municipal organizations interact with the citizens ‑‑ via electronic means.  As well as a third important aspect is maturity of ICT industry, both in private and the governmental ICT industry, how they are able to use possibilities of new technologies, as well big data, artificial intelligent technologies and how the industry develops.

Next, please.

The latest ebenchmark surveys in Europe shows the position from different EU countries concerning E‑governance and during this analyzed, was identified that it's not enough to just make electronically existing procedures from the ‑‑ the transfer from the paper to the digital means.  It means absolutely new approach providing using existing data existing knowledge from the government to make a more effective service for citizens.

So the second aspect is the policy at least in my country that public service must be as much small, and as much cheap, as much effective.  So means that we have to decrease the number of people.  We have to decrease front office workers and we need a new challenges and new technologies to provide the same level, and better level with a smaller amount of persons.

Next, please.

So next level for public service, we see it's ecosystem and new approaches and techniques, which can be used and the first one is a proactive delivery, customer care will move from today's one size fits all model, to an increasingly customized one based and organizational design and specific customer preferred.

So we put the customer in the center of all interaction with the governmental services.  The best contact with the government is no contact.  So the government approaches the citizen, provides necessary service.  Sometimes without any interaction from the citizen, and requesting any services.

Human and machine integration is very important and we see there is some historical background where public service providers, public servants sometimes need to be challenged to be more innovative and attractive and exploring new technologies.

Of course, user driven service design and co‑creation needs more and more involvement of citizens in specific public service deliveries and designing public services.  And how can we do agile for the public services innovative approach.  Sometimes we have to understand that the public service and designing the services can make mistakes and then we have to acknowledge that, but we have to move on and provide state by state‑based approach, piloting projects for artificial intelligence and technologies are used for public services.

And innovation as a function must be implemented in the heads of the public servants and the people responsible for delivery and improvement of public services.  How the artificial intelligence can help.  There are a lot of added values.  How can we improve our services from the government to the citizens.

It's availability of the services.  It's multilingual to serve citizens and other parts of society, I mean in broader context of European Union and global scale, and cross border services which needs to serve not only in the main languages but also the small languages are also needed services to be available for these people in their native languages.

Of course, speed of response time for the services, we don't need hours or days for response of citizens requests.

Speaking about artificial intelligence, and, for example, Google alpha IQ was after the researchers from China, they identified that existing IQ for this program is 4728 points and if we compare to the 6‑year‑old child is 55.5, this means that artificial intelligence is growing and ability of artificial intelligence technology is enormous and for example, in 40 days this program taught herself and improve ‑‑ and become the smartest player of the go game.

So we always speak about employment, unemployment and how the technology will replace or help us provide services to the public and if we compare 2017, the first year and proportion between the caring professions and the power workers, dramatically changed.  And we see the digital care channels the share will grow and the probability of 77% of repetitive and predictable roles will be ultimate mated.  We don't need only the replacement policy, but especially in the government or some services like investigation medicine, but also approach and technologies relieve, split and augment should be explored to make as much as possible and as much as effective artificial intelligence using public service.

And according to Deloitte and Gartner Research, the artificial intelligence will create more jobs than destroy.  Of course, we need a new level of training.  We need a new level of governmental interaction with possible retraining of citizens.

So putting boats in service of e‑Government we see main areas where we can go further with the development.  One is machine translation technology.  It integrates the different public services as well as translated text documents and website with accurate ‑‑ with accuracy, speed, and quality.

The second one is virtual assistance platforms, which is developing very fast and developing the virtual assistant platforms, we see that the customer support search assistant and different other technologies can be functionalities can be used.

So virtual assistance provides some kind of humanity for machines, with social and emotional intelligence, speaking to the customers, but the background of this nice virtual assistant platform is a big and huge, work done by the machines.

So from the practice or main enterprise register company, which is responsible for registering all private as well governmental institutions, started the project that customers will be served and supported by virtual assistance and taking into account that approximately 40% of different company registration procedures are done only in the ‑‑ only in digital means, that means entrepreneurs don't need any more come to the office.  They do everything via Internet, the next step of office and quality, the virtual assistant will provide ‑‑ provide this opportunity.

So what the government can do, in Latvia, we developed so‑called nationally driven concept, which will be in detail described by my colleague and later presentation.  But in general, what data from the government, from the different governmental agencies can be used so providing more accessible and more friendly public services, the innovative how to respond to the requests of the citizens, using different parts of databases, which is available to the government and, of course, assisting in everyday feeds of citizens to set up necessary procedures.

So a data driven is built on three main pillars, the government, the cooperation between private and the governmental bodies, as well as innovative pillar where all of these first two pillars with the citizens engagement and the public ‑‑ the public opening data will create innovative products.

So how can we use technologies and methods to create the best results in public service?  And this is a McKenzie framework of how government can use state‑by‑state identifying the best practice and ending with the open culture and organization.  It can be supported by the practice, for example, for the use case, where we find the best possible player on the governmental side, for example, this enterprise register.  We took this company as a ‑‑ as an example, in cooperation with ‑‑ or private company Tilde.  They had a project to develop this virtual assistant.

Then the data ecosystem needed and data echo system requires cooperation and accessibility of data and as an example we can mention this data driven concept is an ecosystem for this data, availability of data and how we ‑‑ this data will be used more effectively, than technology and tools.  That no how can we ‑‑ how can we build this ecosystem for cooperation between partners from the government, academia and industry.

So we signed the cooperation agreement between my ministry which is responsible for governmental issues and the industry represented by the cessation, as well as academia.

And full work integration, we integrate those activities and specific organization, artificial intelligence is widely used for forestry, as well for cart register in Latvia.  So some practical examples.  And summarizing everything together, we build ecosystem and open culture for society, and society we understand government, public sector, as well private sector and industry.

So a future, near future for application of artificial intelligence, we are looking for regional development, different aspects, how can we use artificial intelligence for regional planning, of course education, and different tests can be applied for different technologies, public procurement control systems, as well as transportation organization systems is already used and my colleague will tell more in detail.

And, of course, justice and court procedures which is very sensitive, but some processes can be delivered also for machines.

So it reads machines are our friends and I hope we will manage that for the near and the far future.  So be trusting technologies and we provide technologies to public service and public service improves the performance for our citizens.

And at the end, we have home page and if you like Latvia, we of course like you.

Thank you very much.

>> MODERATOR: Thank you to Mr. Edmunds Belskis.  So let me pass this to Dr. Signe Balina.  She's an advisor to the minister of ICT matters and associate professor in the university of Latvia and one of the real gems, a lady in ICT.  So Ms. Signe Balina, she will give us information on how to connect the dots and how to foster that debate innovation.

Please.

>> SIGNE BALINA: Thank you.  Colleagues, my great pleasure to be here with you and share the vision as Latvia.  I think the academia and the state institutions, and the aim was to understand the current status, what are the strength of the industry?  What are the strengths of our researchers and the second one, the second aim, what is are the future tendencies on technological development owl the rounds evolved and what can we do in Latvia.  What are our strengths?

There was a good cooperation between academia, but the next question was how it is this corporation among industry and academia players who are interested in cooperation of the state and what is the vision of the state?

We realized all together the data has become the new oil as new currency and there is a need to be transformed, the data of income, the other smart hospitalization industries and the state, of course.  We started to go and still they are working on auditing amount of information and data that the state owns as publicly available and closed and is there any demand for one or another when they intersect.

Finally, Latvia will be the data driven nation.  And in December of last year, Latvia and the government signed a cooperation memorandum, with the aspiration for Latin American via to become a data driven nation and we agreed upon the concept of data driven goals, and three pillars and the common goals in the process of digital transformation of Latvia and development of the company is called.

The major goals of memorandum of cooperation are to strengthen the development of the Latvia industry and export, to use opportunities provided by the industry for the growth of Latvian economy, as well as to achieve full potential of digital government, according to the requirements of modern society and economics and also, of course, to increase competitiveness of the country.  And we base our strategy on the unique moment and the opportunities that enforced the Industrial Revolution provides for a small, start and flexible country like Latin American via.

And the strict pillars are the first, of data democracy and accessibility.  It's a basis for a more transparent public ordinance process, for more active public participation and decision making process for the development of new innovative business models and for the higher quality of public sector and loss business decision making processes.

The second pillar is data enabled citizen engagement.  It's very important that access through transparent and perceptive information about the public sector, activities and results, and it is something like foundation for civic participation and the citizen engagement in the public decision making processes.

And the third pillar, that is very important for the industry is data driven innovations and innovative data commitments and the development tenancies and opportunities.  And this data driven innovation, we look to this through an instrument to strengthen companies competitiveness and the expert capacity of high value added products.

So ‑‑ so government decided to step in and we worked all together.  Government, industry, and academia for professional benefits for all of us.  And Latvia, we are looking to grow the industry and it's designed to create new innovative products and services that can be used in private sector and also in public sector and they are also looking as also Mr. Belskis told in his presentation that we are looking to public egovernance projects that is ‑‑ the innovation platform and the launching pad for innovative technologies.  Projects with high, high commercialization and potential in the future.

There's a capacity of Latvian signs of the ICT sector and competitiveness will be stimulated in line with the global technology development tendency such as big data, and other areas.  Where Latvia has already demonstrated strong capabilities.

There's new development opportunities in other industry and in order to make it more evolving, we as an industry work together with different state institutions and ministries including, also ministry of economics and I think the industry are focusing in cooperation with other smart organizations to innovate in Latvia, and government supports entrepreneurs and they help them test their products.  The four state organizations become as a launch pad and we help to test the virtual librarian so that they can establish these eservices.

In Latvia, the innovation platform, is a framework for new innovative developments and we tested the idea of our platform last summer when the old industry academia and government visited Microsoft headquarter in Redmond in the state.  The vision was made that Latvia will be the first in Baltic States innovation metrics center.  The innovation center is between the university of Latvia and Microsoft and I think the industry and innovations are really important in Latvia.  Back, please.  Back, please.

Therefore, it's opened by president and Latvia, you can see in this picture, this September and he said that it is impossible to be part of labor market if you are not educated in ICT and both economic exchanging all traditional economic branch is being we placed with new ones connected with ICT and its usage.

The mission of our innovation center is to ‑‑ yeah, next please.  Is to enable and promote digital transformation and development of innovative solutions in Latvia.  But establishing an open space where industry professionals, public sector representatives, researchers and the entrepreneurs and of course also experts from the industry can engage and work together.  And the digital transformation, it is not something like meet.  This is reality and they need to work together to be competitive and mark ourselves on the global economic map.

The main idea of innovation center is share instead of one man effort.  And innovation center, science is accessible for business and helps solving society matters.  And all new trends, machine learning, big data, artificial intelligence, exit are from the research labs, and the democratization always has to use the best of the results from our research efforts to put this in the business solutions.

Pear and all of us have our specific education, life and work experience, whole base and knowledge and gained from social groups and engagement.  We all are experts of certain topics and what makes co‑working so productive.

Next, please.

Innovation center is a workplace to engage the decision makers, students, developers, to engage and think together for the best solutions.  Innovation center is a place to validate our ideas and search for the best technological solutions and share the results.

My colleagues, you can see them here and you can see also in the picture, my colleagues will continue about innovative ideas that their companies created and developed in cooperation with researchers.

I would like to say thank you.  This is our drive into the future.  Thank you.

(Applause).

>> MODERATOR: Thank you to Dr. Signe Balina.  Now we will go to the more practical part of our session, and our next speaker will be Aigars Juandalderis.  He has more than 18 years of experience in the ICT industry.  And he's exploring new projects and capabilities and keeping hands on attitude.  He's still an active developer and interested in aspects of machine learning, what classical crypto crafty.  And he's talking about efficient public service.  The floor is yours.

>> AIGARS JAUNDALDERIS: Thank you, Gerdes.

It's necessary and it's required but it's not the end goal, obviously.  It's a great honor for me to tell you what's been cooking within this framework, this hot bed of innovation that we created in Latvia.  Some of these examples you will see already in production, some are in pilot phases are and some are active discussion phase, but all ever them address at least, I would say a few roles that AI can play in the public service space.

Next, please.

Next, please.  And next, please.

So on one hand side, it's automation, and Edmunds has mentioned, yes, snail can travel 100 miles but obviously Formula 1 or NASCAR for our friends across the ocean, are much better means to achieve the goal of transporting yourself across.

And the other area, the way we look in Latin American via at AI is hyperpersonalization.  If you remember all good CRT screens in front of you with really nice exchange for 600 resolution and basically, you are seeing some blop and all you could do, you know, what you saw an you could do just as ‑‑ that much.

Versus two 8k high resolution where you can see every single digital pixel to address each of them individually.

Next, please.

So one area in very special for Latvia is AI enabled identity.  Latvia is unique in terms that Latvia is creating the database and the whole idea is that I am who I am, not because I carry some document or some ‑‑ basically any factor.  I am ‑‑ we think that physical IDs and any form factors are outdated basically remnants of all centuries and the whole idea is to create, to enable citizens to receive services regardless of any form factor, physical form factor they previously had to carry around, like in the form of ID card or SmartPhone with some app on it, et cetera.

Next, please.

Pane and just taking some of the areas we have been working on in that regard is evolving and here I would specifically want to address, e, not I.  Driver's license, a lot of countries are working on replacing driver's licenses with SmartPhone app, but why would we need any phone factor at the end of the day?

Health, today in Latvian hospitals if a person is admitted in emergency room, this person can be used by using biometrics and AI behind the scenes.  The public services where you have to present passports or any form of physical identification.

Next, please.

The other area or application, rather, I would say is transportation.  Transportation is a blood line of any modern economy and, obviously, the ‑‑ sometimes the controversial goals of efficiency, safety, security and mobility, pose quite the challenge for many countries and with the question how to balance ball of these goals simultaneously.  And, again to hint you some of the ‑‑ next, please.

Some of the areas how we are working in real life on the streets of Latvia, using AI, is to attract for those with special needs and have public transport priority lanes and, of course, personalized speed limits for specific context like taxis or buses, school buses in a way that was previously impossible for vendors of industrial sensors of old.

AI today can be present entirely different possibilities for city planning safety and security of transportation.  And you probably saw work beneficial at the title of my slide and I would like to show you a short video from a project in Latvia, which is really in production ‑‑ I will later tell you why this represents in my opinion, this whole ‑‑ it solved the whole idea of beneficial AI.

Next, please.

There should be audio as well.

>> Every year, more than a million people die worldwide because of auto accidents.  That's equal to half the population of Latvia.  In Latvia alone, there are around 4,000 auto accidents and 157 deaths yearly.  The top reason for these deaths, speeding.

This is Joyce who enjoys going fast but today she was running late.

What if we could turn back time and save her?  Since 2016, next generation speed cameras with future intelligent transport systems have been installed throughout Latvia.

The cameras detect immediate limit violations and capture both photo and video evidence at the offending vehicle.  From this moment, the Microsoft cloud and the built‑in machine learning magic begins.  First, an intelligent download mechanism sends the data to the cloud and then a proprietary deep neural network recognizes the vehicle type and its number plate.

Afterwards, the information is preprocessed and sent off to a law enforcement agency.  As a result, the entire workflow is fully automated.  Thanks to Microsoft Azure this process not only requires little human intervention but also shortens the process that once took hours down to within a few seconds.

But most importantly this solution contributes to saving lives.  Statistics prove that since their installation, there's a 42% increase in serious traffic accidents, a 21% decrease in fatalities and less incidents where the speed cameras have been placed.

Drive safe.  Your future depends on it.

>> AIGARS JAUNDALDERIS: And I can share a little secret.  This video is a bit dated already.  You probably saw some company names on it, and Cigna said to visit the company in the summer.  We gathered the latest statistics and what we are proud of that since the installation a year and a half ago, it's still zero speeding‑related deaths at the hotspots where these cameras are installed and it's 70, 70% reduction in serious accidents in these hotspots.

You may also wonder why speed cameras, we have all seen them everywhere around the world, basically, but as I mentioned, the unprecedented level of hyperpersonalization AI brings to us, allows us to significantly contextualize different types of vehicles and traffic flows, as I mentioned taxis, buses, school buses, trucks previously, simply that was impossible to address.

Next, please.

But I also totally believe personally and I must be honest, we don't have a silver bullet in answer to all of these questions but as AI matures, I believe we need to start thinking about how to apply AI in a truly beneficial way and I would encourage you to see by yourself on the web, on the link mentioned on the slide deck these principles of beneficial AI and how they could apply to every single project up undertake eventually because it's far from simple to answer all these challenges in everyday applications.  That certainly is what on our priority list and I hope that going forward we will build truly beneficial AI not only in Latvia but across the globe.

Thank you.

>> MODERATOR: Thank you, Aigars.  And least but not last, I pass it to Mr. Andrejs Vasiljevs, he will speak about art visual intelligent power of languages which can be used and are used to provide public services.  So please.

>> ANDREJS VASILJEVS: Good morning.  Language is essential part of our identity of our intelligence and human computer interaction.  And ‑‑ and this is why we focus on bringing language support and technologies and using the latest advancements in AI to make computers language savvy and language intelligent and we work on the techniques on machine, and natural language processing services and virtual assistance and chat bots that can communicate with people across language manner and as close to human ‑‑ human interaction as possible in dealing with new services and devices.

Next, please.  Our company is part of this active cooperation between private sector, academia and public sector.  We are located in all three Baltic countries, with headquarter in Riga and active part of research and industry collaboration enabling languages with the latest advancements in technologies and AI development.

And why it is so important?  Because in Latvia, they have in the crossroads of economic and linguistic crossroads and being a small language spoken by about 2 million people only, Latin language was able to survive for millennia.  And it's partly thanks to adoption of new technologies for Latvian language.  And ‑‑ yes, next please.

And we believe it exemplifies the opportunities in Europe and globally.  And Europe is very rich in languages.  It's 24 ages and 80 languages are widely spoken by minority communities, regional languages and for one side people sometimes say all right, it's a big problem and challenge what can we do about that?  But actually, we had to look to this diversity and that the problem, but it's an asset.  It's a source of our cultural richness.

And language is a way how we encapsulate human intelligence and language diversity, many dimensioned how to encapsulate that intelligence.

And the question is how to extract this intelligence from the language information and how to enable humans to communicate cross language borders and we believe in Latvia, we have some examples that we can show to the rest of the world.  How to bring advancements in big data in artificial intelligence how to put together many players so that this linguists in this country, there are 7,000 languages spoken that could enable all communities to benefit from technology ‑‑ technological advancements and now to some practical examples.

And in particular, machine translation is very efficient tool to cross the language barriers and access the multilingual information.  But they can help to simply get access to information and when you need reduction or publication, and when the humans can correct some parts that machines are not so good, and more attrition in translation work.  We use these techniques by providing machine translation as essential element in e‑Government service infrastructure.  This platform is a machine translation service platform that enables other e‑Government services to become multilingual and that provides quality translation in a secure way that's able to translate all kind of formats and online information documents and emails and that's specifically adapted to specifics and terminology of public services in Latvia.

To bring the very latest advancements in MT, I would say not all network techniques is a paradigm change in many areas and many machine translation, in particular and there's a huge, huge, wave of development in this area and we are among the pioneers in applying these techniques.  And during ‑‑ we see our competition of the machine translation players around the world.  We were happy to get the best result in machine translation for our machine translation system, beating other competitors like Google and Microsoft and many research groups across the world.

And this is because was to deal with linguistic complexities and we are able to cover complexities similar features could be found in other languages as well.

And that was exemplified, again, by the request of EU presidency and the European Union, as a particular country in charge of the EU, and that country, and with the second half of the year, 2017, Estonia was in charge of the presidency.  And Estonia was smaller language country and in some ways more complex.  We have to deal with language barriers and we help to apply not MT with other techniques and develop machine translation system for EU presidency in Estonia.

Next, please.

That helped both journalists to access local information, delegates to communicate these locals to ‑‑ the local staff to deliver information as quick as possible and professional translators to be much more efficient in where ‑‑ the productivity of our work.

Next, please.  It was also interesting that it brought attention to the top level politicians who gathered that in this case the digital summit in Europe and the ‑‑ sometimes politicians tend to access the things about how we are, if we do not see any real solutions.  And hear some of the politicians for the first time were able to see how AI can be immediately applied to very practical challenges and in this picture, you see how the Prime Minister of Greece, Mr. Cyprus is learning about machine translation capabilities.

And now the next EU council presidency will be in Bulgaria and also we will follow and we will also develop MT for these countries as well.

Another area of machine translation ‑‑ of artificial intelligence, that we actively develop in Latvia, a virtual assistance which was mentioned by Mr. Belskis.  They would have different personalities and know‑how, but to speed up this development, it's important to have the basic intelligence, the baseline platform, that enables the basic functions.  Latvia is very important, that they are able to community in local languages, and we have developers which will assist in the platform on which our most popular assistant Flora is based.  She spoke with 300,000 people already and with social media platforms and also mobile labs and now we put those assistants in service for e‑Government to help to realize this vision of data driven nation to extract power of data, of open data, public data, of how people can ask questions and get answers from virtual assistant in a similar way, how we can approach public service personnel.

And that helps both to unleash the power of data, help people in their everyday needs but builds a different type of connection to services, very ‑‑ they become more personalized and also it creates some kind of emotional connection because these assistants are friendly, adaptive and personalized it and always listen to what you want to say.  Maybe not always able to respond to your requests but try I to do my best.  Another example of application of virtual assistance is in libraries, and we are particularly proud about national library of Latvia, which millions and millions of publications and records and virtual assistance shows how to, again, extract power of that information and help people to find the valuable publications and information which we need in library by simply chatting with virtual assistant in the same way.  How we do that is librarian.

So these were examples how we make ‑‑ use AI to make it more inclusive and multilingual for the benefits of more inclusive services and more inclusive world.

Thank you.

(Applause).

>> MODERATOR: Thank you.  So with that, we conclude the presentation part of the session and during panel, we covered aspects of AI in public service delivery from a government perspective.  So what we learned from that is if we treat properly the AI, it's not a threat but an opportunity for government, for different challenges and that replacement is not only strategy that is most widely discussed when we speak about government and AI applications.

Then we spoke about the multi‑stakeholder cooperation between academia, industry and government that cooperation, collaboration is essential to drive innovation in public sector and as well it could serve the beneficial the other way, the industry, to create innovative products and then we followed with practical use cases of AI applications in public service.  We spoke about hyper personalization, the AID is not dependent on documents and finally the multilingual and public services.  We have million time for questions as well.  If they are in the audience, are there some questions addressed to ‑‑ yes?

Okay.

>> PARTICIPANT: Yes, this is Karine Perset.  I had a question about the speeding application that sent the speed raters that sent data about the car owner and license plate through the private cloud and I was wondering whether concerns had been voiced among the local population about the private company having access to this personal data and information about infringements of the law.  So I was wondering about that.

Thank you.

>> AIGARS JAUNDALDERIS: Yes, definitely privacy a big concern.  ICT in general, so, yes, the solution deployed to operate also compliant to GDPR and new upcoming regulation and yes, there has been a lot of debate around how data in general should be treated within public services.

Speaking about particular case, though, there's no data open vehicle owners traveling within this environment.  It's something that legal law enforcement agencies cover after they get this preprocess data from the private cloud.  So basically, the only data that travels through this infrastructure is what you can see normally on the street, licensed plate, car, what you see on the public street.  There's no personal data in terms of vehicle owners, et cetera, et cetera, et cetera.  And even the decision whether or not it's speeding, it's made by law enforcement agencies when they received data.  That determines it somehow limits the exposure to this whole consent.  But, yes, in general, security, privacy, in terms of AI, and in the age of hyperpersonalization, big data is very key concern and intrinsic part of AI movement, I mentioned in my slide deck, which I'm afraid nobody has a complete answer yet.  It's still an evolving field, not only in terms of technology, but also in terms of legal aspect of this.  So, yes.

>> MODERATOR: Okay.  Thank you.  We can take one more brief question, if there is such.  Okay.  Yes.  Please.

>> PARTICIPANT: Thank you very much.  Sebastien from France.  First of all, I think I am very concerned that if the IGF has a classroom‑style meet, it's very bad.  We can't have a real interaction with the participants.  It's not what IGF is meant for.

The second point is that do you really think that it's with using a stick that you will help the end user and especially the driver here in the presentation to adopt AI or not to fear AI?  I am really concerned that the way the government is trying to AI, it's to get more money or to get the stick against the drivers.

There are plenty of solutions to help the driver not to go higher than the speed limit in their car, in relation to between the car and the different component of the road system, but the only thing you show us here, it's how you will punish the driver and if I am punished, I will not want to use AI and I will struggle against AI.

Then please think about that.  You have user and they don't like just the sticks.  They like to be helped in what they do and not the reverse.

And I am really concerned about that.  Thank you.

>> MODERATOR: I will try to comment about the stake from the public sector side.  If you look to the governments and the governments have a different function, including control and inspection function for different areas.  And in that case, if artificial intelligence can help provide necessary functions in more effective and user friendly way, that is only ‑‑ we can be proud of that.

Of course the interaction of the society must be based on a win/win situation and with proactive moment.

So we have different ‑‑ different applications, for example, with the machine translation and mutual assistance.  There are different ways how artificial intelligence can be exposed in the public sector, even for investigators, even in jails, everywhere can be used and different public functions can be replaced.

So, yes, sometimes we need to use this technology, also for inspection and control functions.  And.

>> With a purpose, we chose this controversial example, this carrot and stick example that you mentioned.

Though the same engine, the same technology can be used not only for punishing some action of somebody bunt also to support city planning traffic flow optimization and also the example I mentioned for people with special needs.  Their mobility requirements.  So ‑‑ yes, it's controversial.  The specific example we mentioned here today, but, no, it's not the only application of AI technology.  This is just one out of context example mentioned and we are happy to share ‑‑ we are sorry, really for the short time, obviously.  The first issue, you mentioned, but we definitely can continue this discussion.  We can finish up and share more details about the thing.

>> MODERATOR: Okay.  So thank you.  Thank you for questions.  Thank you, panelists.  We conclude this session and the other panelists are available for discussion of the session if you have some personal or in‑depth questions.  So you are welcome to approach and discuss them personally.  Thank you and good‑bye.

(Applause).

(End of session)