 Welcome to day three and welcome to this session of Inter intervention points and opportunities for collaboration It's an ambitious session because during this session we are going to try an attempt to summarize remind you of something about day one and also summarize day two which was a super intense super long and We will provide you an attempt of a map and a few selection of points in day two and then after that I'm going to introduce you the five panellists and This is going I'm going to work in the following way They're going to stand to sit there sitting right now in first row They're going to stand and the react to the initial attempt to map day one and day two and provide your contribution It's a very diverse group of people coming from NGOs from academia from Activists and from foundations and from companies. So look forward to hearing from them Then after their short interventions I will open up to you and the survivors of last night that will provide your own ideas and comment and questions because We are at a very interesting moment in these three days because we already have put so much on the table It's a wonderful opportunity to integrate so provide your own point of view to pose questions Of course to the panelists in general to the audience and then we will close the session and gain going back to the five panelists for a short Reaction, maybe you'll pose the questions to to them or maybe they want to add something to what we heard in this Approximately one hour of the session. So let's get started. Please the slides Okay Okay, this slide was already shown to you yesterday morning by Ours and it was so it I will just show it to you without going into detail You already seen it But it just may be helpful to have a quick reminder of what we discussed in day one So something from the key notes and something from the discussions during the day that informed what happened in the following day which is yesterday and After this reminder, let me try to provide you with an attempt to Capture at least some of the amazing discussions of yesterday So consider the following slides of course as a selection is not exhaustive Otherwise, it would have been like 120 slides at least so it's a selection and it tried to provides a map of the discussions yesterday so the first part is To notice that different types of solutions and different types of problems Sometimes we could cluster them by concept very abstract concepts like Transparency access bias, etc. And you heard that a lot during the day in other cases maybe less frequently But in some other cases technologists like me we're talking about specific Technical issues like specific kind of data's or specific algorithms and that's another way of clustering the discussions yesterday and Finally we heard quite a bit and that's another way of clustering the solutions and the ideas Different moments different stages in the process Some discussions were focusing on contribution focusing a lot on the inputs for instance of the process other on the outputs And so that's another way of clustering the second map that I would like to share with you is that We were handling and discussing very different time scales in some cases people were saying It's already late now. We have to act immediately because otherwise is too late in other cases We were saying especially academics like me. We're saying wait a second in order to provide thoughtful answers and thoughtful evidence we need time in some cases years and Both we need both and so the many conversations can be analyzed the look in the time scales of the discussion and one specific issue that brings the topic of time scales is the Analogy with the global movement about the environment that we heard yesterday and also in the first day And of course that means time scales which tend to be longer because it's a movement that it's Still not comparable to the environmental movement and also brings another dimension, which is not timescale But it's a geographical global versus local as you see in the last line of this second block Finally in this selection of potential maps and compasses We discussed a lot as is usually the case about Interdisciplinarity It's generally an important topic that comes up often when we discuss of digital issues But I would argue that you we can make a case that in the case of AI Whatever whichever way you want to define it is even more important Because we bring up issues that are clearly outside the purely technical domain like values for instance and therefore And that's take it as my own personal Contribution on my computer engineer and and so I was thinking aloud with other colleagues Maybe in the curricula for computer science and computer engineer We should make a much stronger case for an education in of course a partial education in humanities and social sciences for computer scientists More specifically tied to the problems raised by AI and maybe Mandatory rather than elective as is usually the case in many colleges and universities and then besides Interdisciplinarity the need for dialogue as usual among different stakeholders, which is Something that we've discussed so many times in these years in our domain, but let me also add about different professions Maybe we can make a case that We can remind ourselves that professions of the profound role in our society that has been weakening in the last decades And maybe we can bring back their role different kinds of professions the professors the diplomats the lawyers the doctors etc and therefore we heard for instance talking about diplomats, which is a profession and Saying they were discussing AI without having you know the proper base knowledge to discuss the topic So maybe we can also find think about involving professional organizations and contribute to a more intense debate among professions Now open questions drivers and applications. This is a super arbitrary Selection of the many things that happen yesterday. And so you will see that for instance came up the Maybe the need for a definition of AI. Maybe a legal definition of AI and let me tell you from the technical point of view as a computer scientist Sometimes I had the fear that AI is Stretched to cover almost anything and that of course it's the end of the definition because otherwise if it's everything Then it's nothing so we have to be a bit more careful Then how do we find values in AI context? when the very title of this conference talking about inclusion of course brings up from the very beginning the idea of values and since this conversation was so wonderfully representative of many different points of view and also many different geographical areas of course values have different meanings in different context and One thing that came up and I personally witnessed it in the conversation yesterday is that maybe It's a chimera to think about the biasing Or being completely neutral Some people were saying it's total. It's a total chimera. It's impossible to be unbiased There is always biased and so maybe we should strive rather than the biasing We should strive to make explicit. What are the values? What do we mean by the values? What are our biases? What are our objectives? Third point what is an ethical best business model for big tech What are appropriate remedy and we heard several potential interesting solutions for that and finally the Algorithms transparency of the algorithms auditing of the other algorithms So how do we audit the algorithms like in the case of machine learning techniques that are not the traditional if than else kind of algorithms And so there is active research also in computer science to shed some light on the workings of these algorithm therefore making it possible to audit them and Then the usual all day very old the question of who are the proper auditors? So who is the custodian of the custodians? And then finally who do we want a recipe so in other words So do we want a transparency on the algorithm and that's it? Or do we want also a more comprehensive understanding of the so-called health effect in this metaphor with the with the medicine Or both probably probably we definitely my personal position is that definitely we want both Okay, super selection of potential the questions that came out yesterday Then like in day two like yesterday. We have Assembled a number again non-exhaustive, but I we hope representative of action items that came out yesterday from Creating leading facilitating interdisciplinary coalitions or platforms The creation of data commons that you heard yesterday open source requirements how to sustain and include the participation and the influence of the global south in these discussions and The network of centers potential role in all these et cetera et cetera will not go over all of them But it's just to give you an idea of what we have collected with our not taking and our thinking so Let me underline and remind you Four case studies from the global south and undeserved voices that Struck the attention of the audience. So it may be I think it's definitely worth it to Go through them again. We heard there's a buffer social in Brazil and it's powerful message that algorithm shape reality And then you can actually do something about it in some cases like in the cases that they presented The case of the troubling case of background check information in Iran So to be aware of how a certain technology apparently created for specific apparently knock was objective can have trouble some Uses a weather forecasting a diagnosed prognosis in Uganda and the worries about human and data security in that case and then the fairly well-studied and well-known cases of Disparative service In the case of you, but also Airbnb Okay, and so this let's remind about these cases and let's keep our eyes open and also as research So let's keep studying these cases in order to Making more well-known and understood worldwide Now let me conclude before I turn to our panelists with a few personal thoughts If I may Something I've heard and or something that I thought or a combination of both What is actually new and what is not new is a usual question But I think it's it's particularly important for AI there are Sometimes I was thinking when I was listening to the discussion, okay This specific problem you could have be posed ten years ago exactly in the same way Even though the technology at the time was not called AI So maybe we can I say that not to for being skeptical, but because maybe we can learn maybe it is already methodologies and approaches and studies that can help us to understand Problems that we think are completely new but actually with a slight rephrasing or light a slight reframing Are actually old and also avoiding the extremes avoiding the typical or everything is going to be completely changed and disrupted We already heard that before and also everything is going to be catastrophic And the civilization is about to end. Let's avoid both. We know that are both untrue Which is exactly the the point about avoiding everything will change both in the positive the negative sense Third point As an engineer I was reading reports including the wonderful IDRC draft report that was Given to us a few days before coming to the to the conference. I kept thinking maybe Of there are a lot of forecast, but extremely hazy forecast And maybe we can make a step forward trying to be I have a more precise definition of what is success in Different domains, so let's discuss about specific AI application and then let's define together what is success Is this 10 percent improvement? 20 30 percent improvement and Secondly the timescale This is this success happening in six months or next year or in five years. It makes a huge difference So maybe could be worth it case domain by domain Trying to be a bit to agree on a definition of success and on timescale and then Use it as a reference and to go back in six months or one year and trying to assess what's going on with AI So that's what I mean by thresholds and timing and finally I think you all agree with me that the conversation This in this last couple of days and today In Rio has been unique Let me use this word unique because I attended many meetings that in some sense Potentially similar to this one, but this to me is unique because the the Growth of point of views and experience it that I personally heard I never heard it this way and this is an amazing accomplishment of the organizers And so therefore is spring the question immediately is like I want this to last I don't want it to be an event point in time that leaves beautiful memories. It's already something is already important But I wonder whether and I pose it truly as an open question to you and to the panelists How can we sustain this? Is there a way where we can stay sustained is? Inclusive extraordinary conversation Okay, that's it. That's my the summary and I thank you all the crew for providing me with all the wonderful notes For making this recap and this presentation is also the basis for the point of Departure for the interventions of the panelists and also from you from the audience The panelists will if they want to react to this or add their own point of view or contribution It's completely up to them. So let me turn to them and Let me invite Allison Guild from the research ICT Africa network based in the South Africa and If you want to stand there and give you Thank you Should I oh, it's okay Good morning everyone I am from research ICT Africa, which is a public interest think tank on policy in ICT policy and regulation With a membership across 20 countries in Africa But I'm also based at the University of Cape Town in the Graduate School of Development Policy and practice So some of the points I'm making are really to the Academy and as a critique of the Academy and some of around I'm trying to build a informed evidence base for policy in this area So the first I just I mean obviously there's too much to comment on all of them So I thought what I would just speak about where two areas mainly around the need for serious and genuine transdisciplinary Research and proper engagement with that not just a sort of lip service to us needing to do it and then also to look at the problems of data we have and how that impacts on biasing participation inclusion etc in relation to AI but some of the broad points I want to make really relate to the point you were making about what is new and what isn't new about AI and really the complexity of the environment that we're in the locating of artificial intelligence in the ICT ecosystem and not as something in a sort of linear and we've kind of gone through all these things and we're now At AI but understanding it in the in the ICT ecosystem because really that explains the multiple Systems of multiple levels of governance global and national etc that you have to have in order for the Inclusion of people and throughout throughout the world. So I'm part of that complexity Really requires that we engage critically with other disciplines So I think particularly in the area of ICT for D in the area of technology And often called society. There's not a critical engagement with a lot of the other thing if we're talking about using AI for development We need to critically understand development. We need to critically engage with SDGs not just use them because you know They've been agreed by a whole member states etc. What are the problems? What are the challenges there? Critically, you know engage with the with the language of multilateral agencies because that is our job as the academy is to engage with that critically and In a public interested way in an independent way to cut across some of the very vested interests that exist in the funding of research in this area And in the absence of funding of research in this area And so even things like you know the fourth industrial revolution It's just it's used in a very uncritical sense from a political economy sense. This is just a late phase Capitalism advanced capitalism the contradictions that we're seeing around automation and unemployment is simply the contradictions of that kind of You know economic and political development You know, there's nothing to suggest that there is a fundamental shift either in the technology or the power relations or anything else that Sorry, I've lost my note here that distinguished this from You know from a from an earlier phase of Development Sorry, I'm just going to quickly see if I can find it again. Otherwise. I'll just talk without it See if you lost it here Okay, so The other point I just wanted to make quite quickly was so you know We need to interrogate these issues if we look at things at the assumptions that are behind a lot of the paradigms for AI is actually the Assumptions are that we've got mature regulated markets that people are able to enjoy consumer choice in and that is innovation Whereas in fact if we look at the You know the impact on AI or the effect of AI or the potential impact of AI and I'm less developed countries We see a very different scenario. So the manifestations that are there I'm not suggesting that the manifestations that are spoken about in the fourth industrial revolution aren't there in mature economies But they firstly don't play out in the same way because of the uneven nature of capitalism and the uneven development across the continent the ability to harness the benefits of globalization in undeveloped economies You simply don't get some of the same manifestations. So for example, perhaps the Concerns around automation in underdeveloped countries are not such a problem because we are unindustrialized And so we're not you know, we're not going to lose those kinds of jobs And just coming very quickly to the issues of data. I just want to speak about the complementary needs in the complementary Disciplinary area the complementary multi-stakeholder area, but in the area of data we need complement. We need to use complementary data because Developing countries and African countries that I'm particularly concerned with are marginalized From big databases from the collection of public statistics, etc We need to look at how we could we need to continue to do Conventional and demand side research, which is the only way you can actually identify the points of policy intervention in prepaid mobile markets You cannot get it from the supply side data and you you know You need to actually go and speak to people Senses AI sensors mobile phones Don't actually tell you if that person is a man or a woman if they're poor or if they're rich, etc So we need to do this complementary data and only by increasing the not only the access because I think it's really important that we move from just connectivity to The usage the intensity of use the ability the demand side the capabilities aspects of being able to use that data that we're going to be able to Even you know be represented in the in the big data in the big data analytics So I mean that really raises the need to Nuance notions of marginalization More all marginal people aren't the same so I raised the importance of that was raised around intersectionality the need to Understand marginalization. They're people a little bit marginalized They're people a lot more marginalized that people are never going to come online And how do we address those from a more demand side public interest point of view than the current commercial? You know appropriate returns kind of models that we've got that have provided enormous services in undeveloped countries But will still never be able to meet the Be a company delivered at a cost that the majority of people can afford so we need to nuance that We need to use data in a complimentary way And this is where I think artificial intelligence can be used to supplement weak institutions weak government weak Data gathering in order to plan better for you know development by whatever you mean by that Economic growth etc. But that will require levels of regulation that are very often not present in the kind of internet and Open domain and you know thing so it might require that we regulate algorithms that we get transparency that we get open data And I know this is very problematic Because in many countries our states are repressive our states are not You have functional democracies etc. So the idea of the state Getting more control having access to your data etc. Are of course of major concern but I think in these discussions there tends to be an absence of Understanding or acceptance of the needful some public regulation a role of the state Even if it's only an enabling role for the state to allow the industry to take off and allow these So thank you and I invite the Chimayia rule you can speak from there if you want as you or you can come here As you up to you from the Center for Communication Governance in Delhi You have heard and seen her already before during this conference Thank You Juan Carlos It's it's a pleasure to be here Because we were one of the very early members of the network of centers I'm quickly going to tell you about the hats that I wear just because it's a particular hat that I'm speaking from today Yeah, mixed metaphors though this might be So I'm an assistant professor of law at National Law University Delhi where I set up and have run the Center for Communication Governance since 2012 and I'm going to talk to you today Not as an assistant professor of law and a researcher that is interested in AI But more as a person that has built one institution within India in the global south to deal with the new issue of the internet internet governance and what it means and as a director of a center that very early on got to be a part of this global network of centers and Has seen the potential that it brings to the table and so from that point of view I thought that what I'll do today is not to discuss the substance You've heard quite a lot of it all day and I feel like the appropriate place sometimes to discuss the brass tacks can be the academic conferences Many of which I'm sure that we will share and that we will meet at so I'm going to do this in three parts One is I'm going to tell you what I was expecting when I came here And then I'll tell you what I heard and finally I will give you what you might call my two cents We call it my two pie say in India And so I'm beginning basically with see I was lucky enough that this this is a conference organized entirely by my very dear friends It's the global network of centers meets Berkman Klein Center meets ITS Rio And so I knew that it would be a wonderful Energetic interdisciplinary room full of people that really understood what they were working with But that are also capable of constructive engagement with each other So I was expecting all of that the second thing which I'm sure that many of you noticed is that there is a Special signaling to holding an event like this in the global south in Rio and in Brazil We've been talking about how to make the global south central to big two debates And this is very much one way in which it can be done Now in terms of what I heard the day it's it's been really interesting because most of the time I live in this law and regulatory Theory silo and this conference is brought together not only interdisciplinary academics But also funders people who are actually working with technology people that are in the habit of building Institutions building out programs around the world and I feel like the way in which they have been talking to each other is really interesting and I say this because The takeaways that have emerged involve a lot of people that are sitting in this room So the very first morning the session that I moderated Kathleen brought up something that to my mind was really valuable Which is that she said that if you want to integrate the global south in debates about AI the way to do it is to invest Right and and over time I have seen people expand what we mean by invest yesterday I heard someone make the distinction between the long-term solutions the research the building of expertise and the more immediate solutions Juan Carlos also highlighted this today the need to react immediately, but also the capacity to be able to do so I think that that's that's something that we really want to keep in mind as we figure out what to do here and then Moving on from there my my To pricey is basically so what do we do and the reason that I'm talking about this from and From a more director of a center hat point of view is that I feel like a lot of the pieces to the puzzle They are in this room we all We've worked on this before in the context of internet governance To give you examples when when my center started there wasn't a lot of teaching on internet governance I find that after six years when I'm looking back at what happens when you fund a center in the global south build out Not just academic programs, but also programs that are able to engage with policy You build a pool of people that I am now meeting at conferences I'm meeting my former students working in various institutions Speaking both the language of the global north, which is necessary for engagement at the global level but also from a point of view that understands our local context and so when I say invest I mean Not just in parachuting in expertise But in helping the global south build something out that is consistent with its own point of view And I say this not just in the spirit of diversity is good I know we should include the global south But from the point of view that in global debates and in the framing of norms the global south has always brought in ideas that have been central Central to the building out of human rights. So for example introducing gender in the human rights treaties That was actually an Indian foreign policy representative that proposed that If you see the negotiations for the trips and the WTO similarly You'll see that the global south had made very valuable contributions Yeah, so I would say that definitely invest and and the last one is that I feel like this presents a great opportunity Both in terms of very constructive things that can be done with AI making helping the global south learn much more than just about AI But also in putting our heads together and building out the solution. We're never going to have one sort of ready Ready to go today, but we we can put it together. I think over time. Thank you Thank you so much She may now we have Maria pass can Alice the ratios the Dallas in Chile executive director. So Thank you. Good morning everyone So I have the privilege to be the first one that will probably summarize this Wonderful symposium from a different point of view from the point of view of the more Activist organization. I my organization is there to see it Alice as Juan Carlos mentioned Chilean based organization, but we work in Latin America in all issues related with human rights and technology So from that point of view the first one the first thing that I want to stress it's My agreement to this call for the multidisciplinary Interdisciplinary work and particularly to reach the separation that many times happen, unfortunately Between the academic work and the activism work We have our own capacities of research, but many times we are short in the resources and the capabilities in house to conduct really deep research that allows us to have Enough data to be really effective in the advocacy policy work that we want to conduct So I really celebrate this kind of conference in which we have opportunity to have more Conversation to enlarge the network of connection with the academic work in a way that we can try to Envision how we can continue working together for the future to solve this really problematic question about how technology is evolving The second point that I want to make it's like a reaction of what Juan Carlos was presenting in his Own summarize of the main ideas of these days It's related with what is success from the activism point of view that I mentioned at the beginning for me success in the field of a AI is related with Arrived to an agreement a general agreement a general consensus about the urgent need of a full human rights assessment in the decision-making of Moving forward in developing this kind of technology in the decision-making in the design of this technology and in the decision-making of the implementation of this technology and When I am putting out to this I want to go further to just the privacy issues that Usually are on the stage when we talk about Artificial intelligence. This is not just matter about to try to reset the Current rules about privacy or data protection frameworks or how to access to more data about how we are being In some way put in a box for this new technology that is the artificial intelligence But it's more about how we try to find the values to design to to implement this technology That preserve the dignity of human beings. So it could be completely Artificial and completely machine driven the technology But at the end the only purpose that it has is to serve human beings to serve us so we need to reposition that in the Analysis and not only the economical advantage that can have the implementation of artificial intelligence, but in the perspective of making really useful for achieving a better and fulfillment of the human rights In in the same line, I want to say that what is new that was another question that Juan Carlos was presenting to us. What is new? I think that fundamentally new in the artificial intelligence issue It's the the problem of the scale and the problem that it's everywhere potentially. It's a tool that Can in some way affect everyone everywhere and without really knowing that what is going on so But there is also a data implication that go further to the traditional concept of a Data management that we have is not only related to the things that are directly Useful to identify person but a lot of Implication on deriving conclusions about behaviors of our population about put a tag to certain groups that Are being made that have been made in in a way that is totally different on which we Have been done in in the history of humanity So I want to conclude just to how I see the possible solution of For these that of course I'm not gonna find here Probably we will not do it in just a symposium even it's a very valuable Opportunity to do this brainstorming and I think that the key issue is to try to find new methodologies to To build this need of fairness as Juan Carlos was pointing out. There is no possibility of the bias Artificial intelligence, but we can build a way to make it more fair and more reasonable And we need a methodology that is stressed the way in which People can have an opportunity to participate in creating these values. This will be the really Inclusive artificial intelligence and for that I think we need to learn a lot of the movements that were going on before in consumer field in in environmental field as have been mentioned before even in labor Field and From there we can have valuable lessons and how to plug Certain instances of more democratic decisions about the values that are insert in the artificial intelligence But for that we also need to burst the bubble not just being the same people that we are here But communicate this message at large to our fellow citizens in our countries Especially in the global South so they can understand there is an urgent need to pay attention to this now and not leave it for the future Thank you. Thank you so much Now different point of view Vera France Deputy director up in society foundations. I'll give you my mic in a second just presenting you Hi everybody. I'm as Carlos said Vera and I'm working with the open society Information program So I want to pick up on Maria's point on bursting bubbles because I think that is Very good approach and I want to speak about five Fields or movements. I think we need to connect with to make progress on this issue some of them obvious but The first one social justice community I think it was really all all of us were touched by this a baffle social yesterday and Carlos brought it up as well because it really sort of connects and shows where the injustice lies in a very Compelling way, and I think we need many more stories of this type It's not always as easy to explain the bias as it is in this case because it's very directly visible So in some other instances probably many of you are familiar with the work that Julia on when our propublica did She basically demonstrated there is racial bias in the Risk scores for recidivism basically in the United States Judges use a tool to determine the likelihood of reoffending and this is used him to determine bail Early release and sentencing and she showed there is a racial bias And exposed that I think more much and that took a year So it is a lot of work to showcase this bias, but I think much more of this type of work needs to be done And only then I think can we also change the discourse? So I was quite impressed to hear that I think in India, you know most stories are positive if about AI not only Was it for or something that are critical? So I think we need to invest much more into these investigations and they need to be of course into disciplinary So bringing together investigative journalists data scientists researchers like yourself and then also Extract from these examples. What are the policy recommendations? And that's where I think a lot of view come in, you know again in the this Baffle social it's pretty straightforward. It's Search search results need to reflect the population, right? That is the ask And then on some other more complex cases, of course the demands are a bit more complex But again, we have smart policy people in this room and that's where I think some of the efforts need to be focused the second Field or movement Which I know some of you are part of as well is the intellectual property or rather access to knowledge movement So we talked I think Mark Sorman yesterday talked about the data commons and how we need to sort of potentially think along those lines To create sort of data sets that can be used By everybody and not just private companies But also this idea of patent pools and how can we sort of take this idea and use it in this space? So a patent pool basically is a consortium of players a Green to cross license patents and in this case it could be to cross license data, right? And of course the question is what's the incentive for a big player like Google to do this and join such a pool? I think it's probably more attractive to smaller companies, but again, it's an idea that All of you who have worked on IP and a 2k are very familiar with and sort of how what's the applicability of this idea in this field? third Is open government data movement? So I think we also heard in this conference that For example the digital corpus of the European Parliament, and I think the Canadian Parliament is being used for very interesting Project to create voice recognition, etc So again, there's a whole movement out there the open government data movement It's probably now 10 15 years old or so and so it's you know Probably worth having a conversation with them about what are some of the use cases? So what data have they liberated in the past 10 15 years and what are some of the use cases for these data sets? Two more and then I'm done and I'll be very quick with the last two You know a lot of this is about I agree of Allison late capitalism playing out. So Really like confronting the power of the powerful and how do we do this? And I think what one other interesting tool is antitrust In we've recently convened one of our partners has convened a meeting of Competition authorities in Latin America to explore how to use the law the competition laws here vis-a-vis these platforms There's quite some action as you know in Europe and in the US There is a seat of hope around antitrust I think especially if the Democrats will get to power in the future again lastly This has come up in a session on info data infrastructure as well This is as I Allison called it late capitalism structural inequalities and I was surprised that Not once did I hear during this whole week? the paradise papers so you know a lot of the powerful AI producers and sort of platforms are Obviously part sort of using tax haven so and again there's a whole Field out there fighting for tax justice and as we're some of these Players using avoiding taxes I think it worth also connecting with that movement around tax justice in how again to confront the power of These huge internet platforms and the future owners of AI Thank you very much Thank you so much Vera and Victor I can one day from IBM Research Africa based in Nairobi if I'm not wrong Thank You Carlos It's been an exciting event. My name is Victor. He wanted I work in the healthcare team at IBM Research Africa in Nairobi and I'll just talk a bit about some of things that I've learned in cost of this event and some of the key Some of the key takeaways and some of the ideas that really struck me for me the biggest pain points when it comes to matters of AI and inclusion is capacity and Education because coming from a develop from a developing country I feel like I strongly believe that we need to deviate from a consumerism approach and move to a point where we're less at the risk of Western imperialist values trump in our own cultures and This is this is very important because it puts us at a fragile position when we need to negotiate things that have to have to do with issues of data governance and and data ownership and also makes us have to fight to encode our own To encode our own ethical values and as my colleague mentioned we need to The global north and everyone really needs to invest and ensure that this capacity is actually Develop and I think that really struck me over the cost of the event is The need for us to debunk the binary narratives And this is important as well in global south because I can speak for Nigeria Policymakers they they don't really the views are shaped largely by the narratives that they read and narratives that they They hear so it's important for experts to put out their objective in my opinion objective skeptical optimism optimistic narratives about beneficial artificial intelligence and This also needs to be done more and more by people from global south thirdly also Also, I would also like to see more and more discussions around the impact and influence of AI on society on in the long term and also in short term and We also need to see how we can encourage more people to be critical about their use of Artificial intelligence. I was speaking with a colleague from Costa Rica and mentioned that young people like they just they just they were readily They readily accept these technologies and they don't really think critically about what the impact of these technologies is We need to encourage more people to think critically about applications of AI in their society and nothing I would like to mention is The need for I mean a lot of a lot of the discussions here have focused have spoken about How important into an interdisciplinary approach is and it's we can't overemphasize that basically we need to expose engineers I come from a technical background. We need to expose engineers to ethics as part of their training I think it's very very important We it's it's it's also helpful as was mentioned yesterday to have a shared understanding of the various interdisciplinary issues Sociologists politicians Technologies we all need to have this common ground Basically, so true in interdisciplinary is key as you mentioned Another thing I'd like to mention is garbage in garbage out as we're familiar with it in computer science The least we can do is to demand that developers of artificial intelligence systems prove that the input to your systems are representative and very inclusive In my opinion, I think the way forward for collaboration in line with this session is the way forward for AI should be guided by collaboration and again We need a multi-stakeholder approach to have meaningful debates about this issue And my final thought is about the global south and the need for us to stop Taking a backseat in my opinion and widen aggressively a participation in AI and I would end with a code by a bram link on where he said Achievement has no color. Thank you Thank you to all the panelists for these thoughtful interesting contributions now We have approximately 10 minutes for the audience, which I'm glad to see has increased in size Slowly coming to back good welcome and About 10 minutes and I recommend that you try to address either questions to the audience or short comments That can help to have a rich conversation, please freezing in here Anything you could do about it would be greatly appreciated something that I find a bit disturbing about this discussion is Something that's missing which is the discussion of autonomy usually when we interact with technologies We have some modicum of control over how we interact with them. I feel like with AI I'm a participant in the Hawthorne experiments. Remember that everyone know what I'm talking about where workers who are observed how productive they were and You know when they worked under light They were more productive and I think I feel that way about my own interaction with AI because I don't know when I'm interacting with AI I mean, obviously now I know I'm interacting with when I use Siri I don't or when I you know use Google search Or when I use, you know, if I did use Lyft or uber But um, I think in general we don't know when we're interacting and so that's why I think the question of autonomy is so important I also feel a little bit that this discussion about Inclusion ignores to some extent the human right. I wish we would think about this to some extent More from a human rights perspective and what I mean by that is just Is this empowering us to meet our potential or is it at times Undermining our ability to meet our potential Thank you. Thank you. Yes over there It's on thank you very much I think I want to thank Chinmai for reminding all of us that in the end this is all about leadership it's all about the Diversity within each one of us that when we wear different hats We also know we attend events under different contexts and capacities And in that case thank the organizers in particular Carlos Leonardo and of course who has an abacement teams for bringing the diversity People we are into this room some of us who come from the right sector if we are left with some of the corporations the kind of Productive discussions and engagements that have taken place in the last three days wouldn't take place I think we start off as antagonists and so when we meet under these environment with this leadership It's really good because we take away lots of Constructive and positive so Chinmai. Thank you for reminding us about that And at the end and after all I want also to appreciate the fact that the network is growing We are talking of inclusion when we met for the first time at the Backman Center. I was the only person who came from who was black and I was worried now When I get so many Kenyans in this room today Kenyans alone and then Africans Inclusion has been achieved to a very large extent and also the global south. It's happening in place in Rio So I think going forward it can only get better So thank you Chinmai for reminding us that we take more aggressive Leadership role in growing the network to make it better for the rest of society. Thank you Thank you Think there is A hand here. I think okay, please remember to introduce yourself at the beginning Hi, I'm Shaz Jameson from tilts in the Netherlands I wanted to raise the question about education and capacity building we've come back to this point a lot about how to help Solutions emerge from the south and be more inclusive globally that we need education and capacity building on a technical level Great, that's not you know, not a discussion but Victor you said that for engineers we need to teach social science and teach ethics and My question is that some of my colleagues at tilt are doing this. They are in Engineering master's degrees trying to teach ethics and it is incredibly hard to say Why is this relevant and to carry across that message about basically if your intention is to build something that will go to Silicon Valley and You know make a million or whatever. Why should we care? And I think I'd like to if anybody has any suggestions. I'm very happy to hear it Particularly about methodology is rather than talking in the abstract of what we need But how to make this Concrete and actually carry it forward. I think that's something we need to still address particularly as a network. Thank you I think that this obviously For me is perhaps one of the most inclusive conferences. I've been to and since I'm mostly Traveling in Asia than for me that was a really great opportunity to engage with the global south I think that this has been achieved and it's pretty remarkable But we still have a really long way to go Because oftentimes we talk about values that everyone in this room agrees with but the minute you bring them out into other countries They're not going to agree with us And we need to bring these disagreements into this room and talk about this because for example I had breakfast this morning with Nagla and we were talking about democratizing I and we said all democratizing I is so great I said yeah, but I have no idea how to translate it into Chinese So the sense that you know We all agree that there are common values that are not shared by a lot of people who are very prominent in Development of AI and I think that another thing that for me was missing is kind of the strategic map of Understanding which countries are going to be very influential and what are the points in which we need to engage with these people and how? I think that there was a lot of Emphasis on building cohesion here, which is fantastic, but we are missing out on a pretty significant part of the world I mean Asia is home to over half of the world's population and we do have India, which is fantastic But it does not represent a lot of the countries in East Asia that do have that progress And I think that this is perhaps a goal that the network could aspire to this is obviously my first time interacting with you so I don't have a lot of information about what happened so far, but I'm Thoroughly impressed and I feel like if you achieve this I can only imagine what you can achieve If you go on and set that as a goal. Thank you. Thank you Yes Thank you a comment to Shaza question on how How to insert social science values in in the education of Technolog of future technology makers and I'm borrowing from the proposals of both pass on inserting values into technologies and vera to to connect with more communities from the perspective of previous examples on internet governance and GM standardization We've been able to explain what was fair use or what were the different values to be respected when we went to public forums of International standardization and it seems that there's not such there are no no such bodies for for the development of of AI or maybe one Carlos or other Computer scientists you would be aware of which I triple e conference is the key global international arena on where to connect not only with the students but with the professors who are currently developing Technologies Thank you one more question So I'll just go really really quickly and Just sorry, this is really loud Echo what Alex said. I just am so honored truly to be here and to have been part of these few days One issue that I think did not come up is actually the the environmental and physical resource toll of AI and Then it's sort of I think you spoke about this yesterday in your talk, but I'm not sure Right like the cloud is is not actually a cloud it takes energy and Minerals from the ground and And that of course is feeding into climate change, which we know affects More vulnerable populations than others So I just think in the future it would be great to sort of tie those concerns into some of some of our conversations Thank you so much if you allow me in the interest of time unless there is a burning question for one of the panelists to go back to them But before going back to them asking them a tweet the long tweet So you can use two hundred and eighty characters if you want But let me react because it's really my field to the comment about engineers and education of engineers and Because I'm an engineer computer engineer And I'm involved those in the rethinking of the curricula in my own university And I think that it's complex because we have to remind first of all the universities are not only for training professionals but also to educating citizens and that has been Largely absent in many European universities is stronger in the US and the second point is that If we remind engineers that they're also professionals as I was mentioned on the beginning the idea that they are Professionals what is a professional? What are the duties of a professional? That's maybe the easiest entry point than to talk about ethics and their role in society I think okay, let's go back to our panelists from Final very quick around the elison, please Thank you very much. I just wanted to Say that it's sort of as if we expect certain things to be happening online when they're not even happening offline So, you know whether it's human rights or whether it's you know connectivity or access we sort of you know expect there To be open data. There's open government whatever it is When it's you know, there's repression or you know, not not a failure to exercise rights So I just wanted to sort of flag that because I think we keep coming back to the same problems because these are fundamental Developmental challenges that we're facing and I just wanted to say in terms of learning from other areas Which I think is absolutely critical and I think everybody's emphasized that is that some of these Intersections require Different kinds of responses. So if we take the commons for example, the commons have really developed In terms of the data commons in terms of creative commons licensing, etc Even copy left in some ways in opposition to Formal systems so they've become formal systems, but they've actually developed in opposition to those Whereas other kinds of commons, which I would also suggest is a way of addressing inequalities and social exclusion and social justice issues In the area for example of infrastructure commons really require State participation and so you've really got to sell the idea of it You know, you can't get open spectrum and open spectrum would allow a whole lot of people who can't get Services through commercial things to access unlicensed spectrum possibly to innovate possibly to create AI, etc But it's going to require even though we can learn in terms of Concepts of openness concepts of commons, etc in application to certain of the challenges we face We might have to approach it, you know differently I mean the open date the open government thing which Kenya was you know a big leader on all Legislations there, but there's literally no open data in practice. So it's actually the implementation That's the real challenge in the developing context Yeah, so so what I've heard is a is a really rich and broad ecosystem of problems And I think that the immediate thing that I would like to see is the beginnings of an ecosystem map So we have foreign policy issues. We have very local issues. We have global north versus global south We have intersectionality intersectionality marginalization academics activists and I feel like really the best that most that any of us can do is find our corner figure out how to be effective Figure out how we can what is the strategic way in which we can join hands to be more effective in particular contexts And then go at it But the trouble with AI is that it's moving so fast that that information is not very organized right now So I think that one of the outcomes that I would like to see from this space and this conference is an ecosystem Mapping so that we can understand how individually and together we can manage to be as effective as possible Yeah, I couldn't agree more with the points that have been already made. I think that Just adding a little bit to the last thing that team I was mentioning. I think that After we have that map we need to be like really strategic in how we use that map in the sense that Not everyone can do everything Each area or each field of work. It's it's different and have like their own advantages So we need to work collaboratively to define the strategies and to allocate the task for each one in who we are better and second and lastly I want to just come back to the idea of this democratizing of the selection of the principles that we want to feed the Artificial intelligence, but for doing that we need to open first the black box because as I heard yesterday in one session We cannot Find what is missing if we don't know what it's inside So we need a the help of regulation the help of The collaboration and conversation with public and private sector to open the box Yeah, so for me I'm wondering if even if we get everything right, so let's assume we you know bake social justice into these systems and Human rights and a lot of the values we care about You know, there will always be things we can't measure Where is the friction that would allow us not to follow rules? Which for me is a lot of you know, what makes us human and I just want to remember some of you have seen it Mimi yesterday Presented couple of artistic artistic project erased Awareness about these issues a self-driving car in a circle So it was in a parking lot and a circle a continuous circle was driven around it And the self-driving cars program to say if there is a continuous line, you don't move You don't cross the line right and so the car was stuck, right as opposed to a human would say, you know what? There's nothing around me. I can drive. So I'm just wondering, you know, even if we get the perfect Sort of rules. Where's the friction not to follow rules? Thank you so I'm just going to add to what everyone has already said by saying that in discussing how We also should discuss how AI would be regulated not just globally but locally Within an accepted framework, right? That that's also inclusive. I think we should individually continue these conversations in our respective communities and strive to collaborate for that in beauty and more inclusive AI society because the benefits are Clear and we just keep we just need to keep having discussions and conversations So thank you to our great panelists all five of them Thank you and thank you all for participating and listening to this session the mic back to Becca. Thank you