WS 356 SUPPORTING LOCAL CONTENT DEVELOPMEN

FINISHED TRANSCRIPT

EIGHTH INTERNET GOVERNANCE FORUM
BALI
BUILDING BRIDGES ‑ ENHANCING MULTI‑STAKEHOLDER COOPERATION FOR GROWTH AND SUSTAINABLE DEVELOPMENT
OCTOBER 22, 2013
8:45 AM
WS 356
SUPPORTING LOCAL CONTENT DEVELOPMENT

The following is the output of the real-time captioning taken during the Eigth Meeting of the IGF, in Bali, Indonesia. 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 session, but should not be treated as an authoritative record.

>> For some reason, the descriptor is incorrect. This is the room for supporting local content development. We hope you'll stay if you're interested in local content development.

>> APARNA SRIDHAR: Good morning, everyone. My name is Aparna Sridhar. Contrary to the transcript, this is on supporting local content development. So I'm hoping we'll be able to have a very robust conversation. We're joined by a number of people who have approached the issue of how to foster and encourage local content development from a variety of different perspectives. I'm going to ask each of the panelists to give a brief introduction. I'll ask a series of initial questions and then I'll open it up to all of you to ask your questions. We'll start with you, Michael.

>> MICHAEL KENDE: Hi. Good morning. My name is Michael Kende. I just recently joined the Internet Society as chief economist.

>> KAREN ROSE: My name is Karen Rose. I'm the senior director of strategic development planning at the Internet Society.

>> APARNA SRIDHAR: I thought I would start with you, Karen. I know you've thought a lot about the relationship between infrastructure and fostering local content. And so if you could just share a couple learnings that you've had at a high level and then I know Michael has done some really specific work that I think probably deserves discussion and exploration as well so I'll turn it over to you.

>> KAREN ROSE: Thank you, Aparna. Thank you for attending this session. It's a pleasure to ask to participate. As Aparna was mentioning, the Internet Society has developed a lot of the Internet structure all along the distribution chain, so when it comes to local access issues, all the way through ISP's and international bandwidth issues. One of the other things we're looking at is the other side of that puzzle, which is the demand drivers.

Not just at ISOC, but on building Internet infrastructure. From the content side there has been a traditional focus of looking at local content from a bit more of a cultural needs perspective and a cultural preservation perspective. There's been a little bit less focus, however, on the development of local content industries. Incentives for local content that has already created locally are aimed at local markets and hosted externally. It's extremely important from the perspective from promoting overall access the Internet and also affordable. Content is what drives people to the Internet. So it is a key part of that whole access ecosystem equation, not just infrastructure itself, but also the information and the bits that are going to fill the pipes, get people online, get businesses online, and get people to use the Internet more effectively.

So while access infrastructure and small scale cultural preservation content creation is important, a greater focus on developing incentives and solutions for creating viable and thriving content industries from content creation to local hosting to subregional contribution is a key part of this puzzle and it's a real pleasure to be on this panel though look at these issues more in depth.

So what is the result of sort of really not looking at the content size and more from an industry perspective? The result is that there's a lot of access infrastructure that's been put in place that may be underutilized, and the broader Internet ecosystem is under developed because you're missing the industrial cultural side.

The other thing, too, is there is a real big impact on jobs and the relationship with the local economy. We know that there's a large base of digital skills already in place in many developing countries. So take, for example, Nollywood, which is Nigeria's Hollywood. They produce over 200 videos or movies per month, primarily using digital technology. And it's an industry that's over about $250 million a year. East Africa is a hotbed of applications in web based in digital services and production skills are booming in places like North Africa, the Middle East, in places like India and other places and Asia.

So these content creators and industries really need to have the conditions to be able to host content locally, closer to their home markets, thereby making the access to their target audiences and their customers potentially quicker and cheaper, as well as the opportunity to host content internationally and serve the lucrative markets which is usually a source of very early Monetization of local content. I've used a couple examples from entertainment, but this also goes to other industries in supporting industries within the economy in general, export, banking, travel. It's also hosted overseas because of a lack of hosting facilities because of the high cost of tariff and other factors, power reliability, poor interconnect activity.

So millions and looking out into the future potentially billions of dollars in jobs, in new knowledge industry creation, and savings and access costs at large.

So I'm going to turn it over to Michael Kende and look at local structure and local content in those opportunities. But overall I think the challenge for us all is to really, in addition to looking at the issue of cultural preservation of content and digitalization of local cultural content, but also looking at local content from an industrial perspective, and what needs to be done across the human, technical, and governance pieces of the puzzle in order to create incentives for local content industries to develop.

>> APARNA SRIDHAR: If you could just highlight a couple of the key learnings that you've worked with ISOC in the last couple years that would be great. Karen really hit on the infrastructure versus content, and I know that's an area of expertise for you; so elaborating more on that would be great.

>> MICHAEL KENDE: Great, thanks, and good local. I want to elaborate a little on the point that Karen made about the local hosting. Start with the observation that I've been doing a fair amount of work in Africa and other emerging regions. One of the observations you can make is all countries ‑‑ these countries all have websites, have content, typically the local newspapers, radio stations might have their own websites, the hotels for getting business, and others, will have websites. But almost uniformly they're all hosted abroad, mostly in Europe. It seems to me that the first step towards creating a content industry that's creating new content is creating an environment in which the existing content comes home and is hosted locally. And that will start the ball rolling towards more people starting to create and put their content online.

So a study I it for ISOC before joining, we looked at the benefit and I am packet of having an Internet exchange point in a country and we looked at exchange points in Kenya and Nigeria. These are first steps for hosting the content locally. So just for background, an IXP Internet exchange point is the place where all the ISP's, the government, education, everyone can meet in the same room using the same switch to exchange traffic with one another locally and prevent having to send the traffic through another country to come back, typically Europe, for instance, a lot of African traffic goes through London or different cities in Europe. A lot of Latin American traffic goes through Miami, probably 80% of it, I think.

So by bringing in IXP locally, allowing everyone to exchange traffic locally, it makes it very efficient to host content locally, because with one connection to an ISP or one connection directly to the IXP, you can get your traffic to all the eyeballs, to all the ISP's in the country. And it will be much quicker to access. The latency is much lower because you're not sending it tromboning across the world, and the ISP's don't have to get it back from Europe or Miami, they can access it on local circuits that are much cheaper to operate.

And Google actually provides a very good example of the benefits of an IXP. Google has a global cache initiative where they'll put their cache of YouTube and other static content that can be stored in a cache in‑country. Typically then through one or more ISP's that connect to the local IXP and make the content available. And it demonstrates very clearly what the benefit of attaching to the IXP is, because as we saw in Kenya, the amount of traffic will triple almost overnight. Once you put in the cache locally, people who are getting frustrated downloading a video from abroad because it took too long or timed out were suddenly starting to watch more video. So they're watching a lot more video. The ISP's are selling a lot more megabits of traffic. So they're getting more revenue. So it really works out for everybody. They're saving costs. It's better for the consumers and they're getting more revenue.

So you'd think that this would lead to a lot more local hosting, having seen this example, but it doesn't seem to be the case. I think that's just because an IXP is necessary but it's not sufficient to promote local hosting. Again, looking at Kenya, there's plenty of data centers where you could put your traffic ‑‑ put your cache or put your server directly, but they're not all full. We have a colleague from the Internet Society and he estimates that probably some of the data centers are just a third full, that they have plenty of space. But they're not full and everyone is hosting abroad. All the biggest websites in Kenya are all hosted in Europe.

So one reason for this and this is kind of an economic explanation, but the reason for this is that the ‑‑ if you're hosting content, it's cheaper to put it in Europe. There's more scale there, better power, big data centers. So you put your content there. In Europe you might save five or $600 a month by putting it abroad. But you're actually imposing a huge cost on ISP's, because someone has to bring it back. It has to come back into the country and the ISP's are paying for that. At 120 or $200 a megabit, you're imposing huge costs on the ISP.

So a solution should be available whereby everyone can save and bring it home. The ISP's would save a lot of money. There would be more page views, and you get overall benefits. And that's hard to come up with, because there's two different players. The ISP's would like to save money by bringing it back, but the owners of the content are saving money by putting it abroad.

One solution you see in a few countries, and I'll just talk about one, we can get into more discussion, I'm sure there's other ideas, but one solution you see in some countries like South Africa is that there's differential pricing for accessing data. If you're using a connection, all your local content is free; all the international content is subject to data caps and pricing. So that means you have even more incentive to put your content locally, because it will be accessed much more because it's free or much cheaper to download. Things like that can help to adjust the pricing issue and bring more content back into the country. That will build upscale. That is one of the best ways to kick start the development of more content is to have the content that exists be brought home, as it were, back into the data centers in the country.

>> APARNA SRIDHAR: Thanks, Michael. I think on this side of the table, we'll move up the proverbial Internet stack. Michael and Karen give us some about infrastructure and content. I'll ask Alice to talk about the Generic Top Level Domain name space. Talk a little bit about the potential that gTLD's have to expand the local content universe.

>> ALICE MUNYUA: Thank you very much. Alice Munyua with the African Union Commission that's responsible of the sponsoring organization for a new Generic Top Level Domain. It's original Top Level Domain. Just going a little back to how the process has been taking place, I can introduce this new process, introduce new details around and launch the process between January and May 2012. The Africa region, having been the only one that actually doesn't have its own Top Level Domain also applied for the geographic Top Level Domain .Africa. It's being sponsored by the 54 African governments.

This expansion, I can new details, is expected to provide a new platform for CT's, geographic, also internationalized domain names amongst people and they intended to allow new Top Level Domain operators to create and provide more content in various languages and also in various scripts. This expansion will allow businesses to identify themselves by sector or by the community. It's allowing the African community and the African region to be able to identify or to have a digital presence, an online presence as the region.

In the new detail process has also introduced a rights protection mechanism or what they're calling a trademark clearinghouse. That is a particular concern to the African region, to us in particular, because it hasn't really ‑‑ as I think as we all know, Africa is the continent that really hasn't had a strong presence online. It comes to intellectual property; we still have a long way to go. Most of us are actually not engaged with this process. That's the reason why the African union is actually through the .Africa process trying to create their own local approach to it. They're calling it the max validation system.

So for .Africa, it is to enhance Africa's digital presence and then bringing the African content together under one umbrella for all activities like e‑commerce to technology and infrastructure to flourish. Out of the Top Level Domain, the African union is going to be creating what they're calling a .Africa foundation that's going to be looking at coordinating ICT‑related activities. And out of ‑‑ it's going to be out of the surplus of managing the TLD. From this surplus we'll be engaging in skill development activities, strengthening Africa top level domains. We know it is an important part of the Internet's ecosystem, and a place to build a strong local community and drive innovation. So that's one of the other activities that the .Africa Foundation is going to be look at, including supporting the introduction of internationalized domain names at that level, which will enable Internet users to access in their own language, generally advances multilingualism, a robust and thriving Internet that also provides local content.

In terms of content development, we acknowledge Africa is a regional heritage, but it also remains completely inaccessible to most of the population. So we call it a content divide. We are going to be looking at utilizing the .Africa Foundation as an opportunity to promote that content development by partnering with content providers, media governments, already part of the government as well as the community.

We look at growing that through providing a space for technology to grow by supporting local registrars. We will be expanding that as well, and also the idea of growing the level of multilingualism at that level.

Thank you.

>> APARNA SRIDHAR: Great. Thank you. Turning over to Zahid, Zahid, we have talked multiple times about sort of your experiences in being a part of the legal and policy ecosystem fostering local content development in Pakistan. So I was hoping that you could sort of offer some high level strategies paced on your experience of what has worked and what hasn't worked in that space in recognizing that every region, country, area, is different. But I think having a diversity of perspectives is really helpful.

>> ZAHID JAMIL: Thanks, Aparna. It's great to see everyone is here at 9:00 a.m. in the morning. This is nearly a packed room. We hear a lot about local, meaning you need to localize. If you want local content, you have to have local ‑‑ you got to have local infrastructure. I come from a different perspective. Is the first step you should build local infrastructure? Is that the driver towards local content being developed? It can be. It's definitely helpful. It's a good thing for that country to have the good infrastructure. Forgive me, I'm a lawyer, that enabling environment that enables this is probably the soft side of it or lawyers might think the legislation is the hard side, but anyway, it needs to be in place so that in especially developing countries people are enabled by those policies to be able to develop this.

So I'm going to talk a little later about what the challenges we have with local hosting in practice in developing countries. Sort of focus on that in a minute. But you need to have access to the platforms internationally. That's my first headline. If I have a local hosting, that's great, but I need to have access to Facebook, YouTube, Yahoo, everything I can get my hands on. So I can learn from it, understand how it functions, use it, and hopefully as a result, innovate and create my own content.

The initial part of creating local content is looking what people have done in other countries and replicating that sometimes. The other part is then to say when you reach a certain level of replicating or trying to look at best practice, saying, all right, how do I make it more local? And you start innovating. That is what drives. To expect someone to immediately innovate immediately without having that kind of access, I think, can be a challenge.

So what do I mean by regular environments? I'll give you some examples that didn't work really well where I'm from in Pakistan. For instance, there was this element that decided to try and shut down Facebook. Facebook.com should be shut. There was a religious issue, and you remember in Indonesia, and the guys that wanted to shut Facebook.com guys were also the guys that had Muslim Facebook.com.

Forgetting about how they went to do and launch that website, one should expect that would thrive, local innovation. You block Facebook.com. This should just sky rocket through the roof. The answer is no. It didn't happen. In fact, that website completely failed. So even if you were gonna try and sort of push other people out, the internationalization out and say we're just going to use local content. It doesn't work.

You have YouTube.com which is currently blocked in Pakistan. What happened to people that were developing contents or upload go videos or linking this to a website? The trouble is when the websites put it up, they don't run YouTube. People starting to say what's the point of knowing this? And so you are seeing less and less videos being uploaded or created for this purpose unless it can be hosted internationally. Now, who can get through to them? You must remember by these sort of regulatory regimes, have you been able to block that? No. Anybody who can get a VPN pretty much gets through. Guess what? There's a cost to it. And what that does is creates a digital divide. It will take a little more bandwidth, at least, if not the cost of having a VPN, unless you've got the free ones running and the ads, and then you think there is a security issue related to free VPN's. What you're doing is not enabling the local content and saying you don't have a competitor from abroad, a big company, look at something local, why don't you develop a local platform? And it doesn't work because it kills the development of local content.

The trouble is, now our government is actually embarrassed. It's not the great sort of being upset with Google or YouTube has run its course. Now it's, oops, how are we going to do do this? It's an embarrassing thing to do. It doesn't work.

I'll talk about hosting and why local hosting can or cannot be healthy necessarily. If you're going to try and have local content developed, think globally to have an open global network and access that is global so that it can have local relevance. So the content is local relevant content. But you need to have open access to global platforms. That is the most important thing, because content does not equal hosting. Content is content. It can be on a hosting that is abroad or local.

Now, let me speak to you ‑‑ we also heard about new gTLD's. It's a global process. And you need to be able to have open access to those.

So what happens when you have local hosting? Well, we've noticed at least where we are. Government servers will tend to run on local servers. The trouble you have is, number one, sometimes they are slow, even though they are local. They tend to be slow. So people trying to access from abroad tend to find there's a delay. It's a reverse. You can't get through the government website.

Secondly, it does cause a huge expense. Sometimes people expect that because it's locally hosted, it will be cheaper. But what we found is, because it's locally hosted, the companies tend to actually take a little more money or at least the same amount of money. That does happen as well, because they are actually trying to monetize the fact that you do not know how to do hosting abroad by yourself. So, okay, we got a guy who doesn't know how to to it. We'll charge him more. This tends to happen a bit. So you know it can actually impact efficiency sometimes as well. I wouldn't say hosting locally is a bad idea. I would say it's a good thing. If it can help infrastructure, it can help investment, it can help having investment networks at home. It's a very good thing. It's important not to say, at least in my view, if you're going to have content that is locally developed, the first thing you need to do is make sure you also have a locally hosted infrastructure. I think that's not necessarily the first thing you want to do felt you want to make sure that the access to the networks are open, that people are able to learn and innovate.

That's my closing. I hope there will be more questions.

>> APARNA SRIDHAR: Thanks Zahid. I want to turn to Bertrand for an overview. We talked a lot about local content. Bertrand often talks about and thinks about is how we sort of take these local issues, cultural norms, etc., and translate them to an Internet that is fundamentally borderless. And so I wonder if you would talk just a little bit about the relationship between local, even niche content, and how a borderless Internet enables or does not enable that content to reach its audience.

>> BERTRAND DE LA CHAPELLE: Apologies for having arrived a little bit the late. Preparatory breakfast. If you're interested in jurisdiction and work ship issues, there is a workshop at eleven.

It's a very good thing to be at the end of the line, because it adds additional layers to the things that I prepared and thought about saying. I like very much what Zahid just introduced. The notion of ‑‑ I wouldn't say the decoupling, but the local content and local infrastructure roughly. Actually, local content, local framework, one thing that is clear. As you said, I'm running this project on the tension between the cross‑border relationship of the Internet and the services that it has, and the patchwork of national jurisdictions. One thing is absolutely obvious in that regard I would follow what Zahid was saying, is that the existence of cross‑border services, sort of infrastructure services in hosting in Cloud and platforms, the best knowns being American platforms, but there are also platforms in other countries. There are very large platforms in China. There are some in Russia. Each of them is expanding in its own cultural environment. It's interesting to see that in a certain way. You see digital territories emerging whereby people share a certain type of script or certain type of cultural environment are sharing cross‑border platforms. It's not only the international western‑based platforms. But one thing that is very important is that the existence of those services, which are sort of software infrastructure, if you know what I mean, it's the distinction between having the hardware hosting centers and the pipes and the infrastructure for the communications. There is a sort of cost infrastructure layer, which is those platforms felt and the existence of those platforms, as indeed produced fascinating effects in terms of lowering the barrier to entry for the distribution of content.

Because when we talk about content

There are many things behind the term content. I'm a little bit sad that this is the direction that it has taken. Now in talking about content, we tend to think in terms of the general intellectual property debate. It's content, like entertainment content. That's not what we're talking about here only. It's also everything that people are posting on their own, like the things that they post on blogs. It can be things on Pinterest or any other social platforms. The reason it is leverage for people individually is that it provides a distribution panel. The beauty of the digital era has nothing to do with which platform you're using. The beauty of the digital era is to lower considerably the cost of distribution, as the music industry, unfortunately for them, knows pretty well.

The replication mechanism -- think about it -- especially when you're talking about a diaspora, when you're talking about local physically, it can be a subset of people. In Asia in the Philippines and a certain number of industries, there is a huge diaspora, and other interaction, they're keeping in touch with people who are thousands of miles away. Is what you're posting local content? You bet it is. It is about the individual life of people, what they think is happening in their country. It's about what they think should happen in their country. The question is not so much that it is the equation is local content, i.e., locally hosted, i.e., locally hosted and so on. Furthermore, this is a whole ecosystem around this. For everything to work well at the local level, you need to have alignment of the creators, whatever the content is, the services hosting platform in terms of the software layer, the hosting platforms in terms of the hardware infrastructure, the legal environment that allows you to post something or not to post something, but also the underlying infrastructure, if there's a power reliability and so on.

As we said before, where in this paradox situation in this content may be content in a distribution network, where meanwhile, very locally produced content to increase accessible to diaspora may be posted in another country. The challenge, though, is what was mentioned earlier. There is a connection between the transmission costs, the hosting costs, and the distribution of value along the chain.

Locating the hosting of content in one country is connected to how much it would cost also to make the back and forth through the transmission lines to a hosting system in another region. And so as in many cases, these issues are interconnected and can only be really addressed in taking the different dimensions and giving the actors around the table the different pieces.

Finally, as you heard, I'm directing this jurisdiction Internet issue. One of the challenges is the compatibility of the norms regarding content in different countries. What is legally in one can be illegal in another. And one of the big challenges we can develop further, but one of the very big challenges is what is the international region that applies to content in one place and may be perceived as illegal in another place? And we're talking about local content. I'm sorry, local content development is not only a question that regards the regions where there is less content that is being produced currently. That's one part of the issue. But there's a big question regarding what is produced in some countries, freedom of expression in one country, that is considered as illegal or offensive in another country.

Here I'm not even talking about religious content. There’s the tension between the United States and France. I am French, and there is a tension. Part of the things that we're trying to address in Internet jurisdiction project is what we call the question of digital coexistence, shared online spacing. What we're talking about is shared online spaces. How do you manage the coexistence of different individual and legal norms? We have a challenge because if we don't find the solution to address those issues of potential tension, then the tensions will arise in the physical world. The whole question of the creation of content is putting people in contact with other people who have completely different framework. Before they were not particularly in contact. It's a great thing. It's wonderful. It's absolutely beautiful. However, it is creating new types of issues. And it's important that we understand that there is an ecosystem of factors at the economic level that contains both the ISP's, the content producers, the hosting platforms, the physical hosting platforms and the software hosting platforms. But there is also an ecosystem of framework of the platforms that host this content across borders, and the legal environments in which each of them operate.

Thank you.

>> APARNA SRIDHAR: Thanks, Bertrand. So we've covered actually, quite a lot of ground, local content, content contribution costs, local hosting and infrastructure, regulatory frameworks, both local and international. We've touched on a lot of sort of areas of expertise, whether they're technological solutions or infrastructure frameworks. We've touched on some economic issues. Bertrand mentioned broader legal considerations and social norms. So it's a lot of ground. I want to turn it over to you all for questions. I have some of my own if you are feeling shy. But it doesn't appear to be that way. Is there another mic?

>> Thank you very much. I really learned a lot through that discussion. It was ‑‑ I'm glad that all of them were short and sweet. It was easy to follow. I have a question that because there are so many other things to discuss which I had never considered, but one thing that I struggle with is language. And I don't know if any of you all deal with that. So I'm more sort of obsessed with the actual content part rather than I hadn't thought a lot of the infrastructure and legal issues that you were talking about, though since I work a lot in the Middle East, North Africa, I do think about the filtering a lot. I was wondering if any of you had encountered this language issue how you dealt we it. Primarily I'm talking about how a lot of populations don't ‑‑ are not comfortable in English, maybe aren't literate in English and access to the Internet is, therefore, blocked to them. That way, as they try to develop local language content, technological or other struggles. That's of interest to me.

>> BERTRAND DE LA CHAPELLE: It's a very important question. I'll make an analogy. If you live in a small town, you have a few shops and they tend to be relatively general. You get a fashion shop, but you don't have 200. Or you get an antique shop and that's all. If you live in a large metropolis, you will get someone that specialized in antique broaches from the early 20th century. Why? Because there's market available. Why is it connected to your question? Because language diasporas or language groups can be geographically and reached a critical mass in terms of outreach that you couldn't in the past. You couldn't have capacity to print a newspaper when your language community is spread in very different geographical locations. With digital space you do have this. Meanwhile, the challenge, and I'm French, so I note problem from the inside, although I'm not an endangered language, it is always a challenge between whether you want to address your own community or reach out to a more global audience. The fact that some languages are gateways to a more global audience is a fact of life. And you see in the scientific community, where we like it or not, that publishing in English is a necessary step. The language issue is going in one direction, which is the capacity to have more reached for less frequent languages, which means that they are likely to survive more. But on the other hand, there's a concentration in some zones.

The second point, quickly, that I want to address is in the new gTLD space, the introduction of IGN is something I have pushed very much when I was in the governmental advisory committee as vice chair, because it is very important to give this differentiation. That being said, when I went to a Russian IGF in months could you two years ago and I was given business cards that were all written in Syriac, and the only reason I could read is the actual e‑mail address in Latin character, I suddenly realized what I had enabled, that the next step will be the business cards entirely in Syriac, and the e‑mail is something I cannot address on my keyboard. It's a very interesting challenge that we are both enabling diversity, but at the same time it may be introducing new separations. And I would even go one step further. I'm on the board. When we think about what kind of top level domains will be accepted in the future rounds in the different scripts, there may very well be different positions by the Governmental Advisory Committee whether word is acceptable in Arabic, but not acceptable written in Latin characters, which means little by little we might emergence of script plates, the rules for the domain names might become slightly different from the rules for other domains in Latin characters. I don't know. So the language issue is becoming a very interesting structuring factor that is not related just to the physical space, but the cyberspace is going to be structure, I believe, more and more around script spaces and the interfacing, both which is very interesting.

>> APARNA SRIDHAR: Alice, do you want to add a couple facts?

>> ALICE MUNYUA: I think Bertrand has said everything. But coming from a country that has 45 languages, a continent that probably has over a thousand languages, the issue of language is very, very important. And not the least the fact that Africa is also divided. We have so there's always an issue there, not just at the local level, but also at the governmental level when we are discussing issues at civil society. It's a sticky issue. It's an issue that is extremely important and that we are always having to think about all the time.

Having said that, though, acknowledging that it's that important, one of the things that most governments to assist in this is to create that environment as Zahid had spoken about in terms of policy and regulation environment that begins to take into consideration some of the local languages as official languages, and ensuring that there is space in both digital level and the way that content is presented and provided and even packaged, that allows for all the local communities to be able to access it, because local content is only local if it's relevant in a language that can be understood by the people that are using it.

But also, we are seeing an increase of more Internet users, for example, in Kenya, beginning to actually localize content, developing content that is local in our own local languages so they are able to speak to users in a more relevant way.

Then an important issue is the internationalized names, that's important, and ensuring that we are unveiling that at that level for governments and for all the other users, so important issue and one that we are taking into consideration.

>> MICHAEL KENDE: Two quick points. Why global is actually local. There are 200 million people. I'll give you an example. We have a language called Urdu. We have a script that is unique to us. Guess what? We're not in the Unicode. The national script doesn't exist in the Unicode and the issue is why? We haven't had engagement in our community ‑‑

>> Zahid, could you explain folks what the Unicode is.

>> ZAHID JAMIL: My fault. I brought it up. Most of the scripts or letters that you use in the computer, I'll make it as basic as possible, are on a table, a big table of many pages. And if you want to use a specific script, whether it's in the Latin, as we say in English sort of characters, or if it's in Arabic character or Chinese character, it needs to be in this table, this book, this big book. And so the thing is if it's not there, you can't use it on the Internet or through your computer. So if my script, which is Urdu doesn't exist in there, I can't even type in my content. Guess what we had to do? We're using Farsi. We are using Farsi for that to be able to type Urdu. The funny thing is we have a smaller province called SYN. It has a totally different script. Its script it in the Unicode because those local community people went globally and said our script is important. We want this in the Unicode. But the government of Pakistan hasn't been able to do a thing about getting its own script in, which is the national official language. Kind of a reverse situation. So that was one point.

Second, think about the issue of if we're going to take content locally to developing countries particularly, it's one thing to have it enabled in the system so up the Unicode, you can have IDN's and all these things work. Also think about are they going to be using mobile devices. If they're going to be using mobile devices, that's the input goes ‑‑ I'm getting all confused here. And does the keypad have that local character on it? So, again, unless the manufacturer develops that with the local character, you have this in Arabic and other languages, it will be difficult for people to be able to use it as well. That's another thing to think about. That's why I think, again, I come back, global access, global impact causes local relevance.

Thank you.

>> APARNA SRIDHAR: Next question?

>> MIKE JENSON: Hi, Mike Jenson from South Africa. I wanted to thank Bertrand for bringing up the ecosystem approach to this. I think that is really relevant. You have to have all the pieces in the puzzle working together for local content to be developed. Building on that, I think for me still is the biggest barrier to local content development in developing countries is the high cost of access. On average this is about 50% of the average developing countries' income compared to about 2% in Euro. So that high cost it is not going to see local content developed in a broad scale and the masses of people to be able to participate in the system. And I think that comes back down to the local infrastructure issue. There's not really a lot matter of content hosting. Of course that's part of the equation, but the provision of access for the end user. And I think part of the problem there is the focus on mobile broadband, which is obviously a lot more of an expensive solution, and it has other impacts on local content development.

People with the tiniest interface on their phone are not going to do much more than hit "like" or "dislike," as opposed to actually putting relevant content on the net.

Also what you're seeing is this reverse net neutrality coming up. We see a few that are free such as Facebook or Wikipedia, or people think they've got the Internet, but they're not being exposed other other sites. All they see is Facebook or Wikipedia. None of the small emerging Facebooks or Wikipedias, are not being seen because it's not visible on their platforms.

Thank you.

>> APARNA SRIDHAR: So maybe in line with that, Michael or container, could you talk a little bit about sort of how ‑‑ what are the strategies to bring down the costs of access more generally, speaking to this more general relationship between content and access.

>> MICHAEL KENDE: Well, that's an excellent question and a lot of work has gone into that. There's a lot of different answers. In some doesn't there's very high taxes on handsets. There's not enough spectrum, so it's quite scarce. You have significant costs for accessing rights of way to connect to the gateways, international connectivity is very expensive. There are a lot of time pieces to the puzzle and a lot of work on that to try to drive it down. The Internet Society is one of the sponsors of the new alliance for affordable Internet that is trying to help meet the broadband commission goals of getting the cost. I think it's under 5% of annual income for broadband. So I think there's 300 per megabit to access abroad. The other parts, anything you can do to take that piece out, has to help to drive down the cost.

>> KAREN ROSE: I wanted to make a comment on what Michael was saying and Zahid as well. ISOC participated with the infrastructure, and one of the key drivers for the local content ecosystem is actually access to international connectivity and international bandwidth. So this key about ensuring that open global access is important is a key piece of this. It is an ecosystem and there needs to be multiple opportunities available for people either to reach audiences internationally, to host things more locally. The key is to put incentives in place for people to have more choice about the content that they create, the audiences that they're trying to reach, and where that content could be accessed most cost efficiently and effectively.

So these are all related issues. Indeed, it's an ecosystem. In general, we kind of tend to think of three very broad buckets, which is the technical infrastructure, the human infrastructure, and the governance infrastructure. These are all key related pieces. When you're talking about access infrastructure in general, whether you're talking about content, and the key relationship between the two, content as a demand driver for local access. And I also take the point of what Mike Jenson said as well. We are selling a lot of the world short when we say it is sufficient for individuals and users to only access content on things like small screens and mobile phones when maybe what you can do is sort of press "like" and type in a few words.

At the same time, in looking at internationalization and international languages, this also has a real impact on the access side of the equation and bandwidth, is the use of audio and video for content. The availability and ability to make a short video and posted it on line, put your music online. These are not inherently text based. They are high bandwidth drivers. So this is all part of the ecosystem as well.

One final comment, one question is what is local content? Right? What is it? How do we grapple with it? I think still the highest number of page views on YouTube is still Gangnam style. It was in Korean and it was put on YouTube, but it's local content to me, because we play it at weddings all the time at and everybody's dancing. It's Korean. To me it's local content. So one of the questions is what is really local content mean? Maybe we're getting a little bit cryptic in the discussion. This is obviously a very rich issue with a lot of interesting angles when we take an ecosystem perspective.

>> APARNA SRIDHAR: I know Zahid had something to add. Then I want to actually turn ‑‑ I'm sorry. Give us a second. Okay. I want to turn to Alice to touch on a couple threads that have gone through the discussion so far. So go ahead, Zahid.

>> ZAHID JAMIL: Thank you. Karen, excellent points. I must say absolutely right, the videos, etc., extremely important. I want to also take this opportunity to say something, which is slightly related with the enabling environment. We think that wiki is done, but let me tell you, this is the question that was the wicket, which is the convention that led to two sides of the world saying we follow this is actually happening on the ground right now. In my country, for instance, we're seeing a reversal of the telecom liberalization that took place in the 2000's. We're seeing Monopolies being created. We're seeing fewer telecom operators. We're seeing fewer licenses. We're seeing a coming together of certain developing country or south to south cooperations going in, leveraging their relationships with the government and saying can we actually close down on this infrastructure a little more? Could there be fewer players so we can help you more to filter? Can we do more? Now effectively from 250 ISP's we're down to literally one or two. That's it. One of the things they're not letting happen at the moment is get the 3 G option off the ground. The point I wanted to make was we're seeing at least some in developing countries a reversal of the liberalization that got us all the stuff that we're using right now.

Maybe one of the things we want to focus on a very high level is how do we get back that liberalization steam. The W 2 did a fantastic job, but we're seeing a reversal. That directly impacts many of the things we're talking about.

>> APARNA SRIDHAR: So at this point I wanted to sort of tie together a couple different threads. Zahid, you sort of mentioned this what do we need to do to shift norms to get back to a place of liberalization and competition? Alice, I know you talked a little bit about capacity building and the work that .Africa Foundation is doing. Karen, you mentioned the importance of not only technical infrastructure, but human infrastructure. Because you are doing some capacity building through .Africa, what I wanted to ask you, Alice, is what are some of the key human capacity building initiatives that we need to pursue if we really want to foster a vibrant content, local content ecosystem?

>> KAREN ROSE: Thank you. That's extremely important. And I think all stakeholders have a role. I'm not just speaking here for African governments and the role they have to play in fostering innovation and creating that innovative environment for content creation, but for all stakeholders. For example, from a .Africa perspective, I expected for African governments and through the African union is having created that environment, provided that environment for just the TLD itself to have been applied for and provided that political support at the national and regional level. But also taking it seriously in terms of beginning to look at, for example, one mechanism we are calling the resolve list and looking at that specifically and connecting that actually to local relevant content and resolving names that eventually might be used for all sorts of things; for example, for education, for tourism, for data creation, even government ‑‑ most of the government's coming up with the open data initiative to just ensure that public‑held data is made available in various languages and in digital format.

For the business community is supporting the growth of the domain name sector in the African region. We don't have that. We don't have that for the various reasons that have been presented before, the digital divide, the content divide, the expensive access, lack of access to infrastructure. As long as we have a very young Internet community and Internet industry, it's going to take a while for, for example, the name sector to also grow into local content to be created.

Also, it's also utilizing the opportunities that come up in terms of international collaboration. There's a level that the expectation is that local governments have to do their work, and so is local businesses, but also at the international level, that collaboration is very important.

Thank you.

>> APARNA SRIDHAR: Karen, is there something you want to add, and then Bertrand, a human capacity building?

>> KAREN ROSE: Yeah. And I think there's also, again, and I gave a few examples of sort of key core skills in adjacent industries that are likely readily adaptable to the creation of Internet content. Again, taking the example of the very developed digital film industries in a lot of regions around the world, whether it is Nigeria, that has an audience across all of Africa and a very large diaspora. I can't even tell you the number of African soap operas I follow on a regular basis, the ability of telling stories well in terms of their voice and perspective and points of view. One of the questions, I think, also from a government perspective in terms of fostering local content skills and industries is what kind of adjacent industries can we look at to say, yes, that are really key adjacent skill sets here that we can incentive eyes in looking of putting more content online that they're already creating or applying their skills to an online industry, as well as key skills to keep infrastructure running as well.

As Zahid was saying in terms of sometimes hosting locally can be a lot less reliable than hosting internationally. There are beyond sort of skill sets involved. There's other issues with power and connectivity and efficiency. But there is a real need to ensure that the human infrastructure is in place in order to be able to support the development of content, whether it's hosted internationally, whether people host it locally, and to really kick start the knowledge economies in some of these areas.

>> BERTRAND DE LA CHAPELLE: I'd like to come back to a few points that have been made. As a matter of fact, when I participate in panels, I always try to pick on something that I didn't visualize as an angle or way to highlight an issue. The connection that was made earlier between VDO, local content, the question of literacy, the fact that VDO is actually something that broadens the capacity of people to access local content because there's not the filter of literacy directed connected to the fact that video is a high bandwidth thing and it's mobile and the broadband network is a key component for the capacity to produce, because the mobile phone and Smartphones may be small, but it produces immediate back loop. We had a photographer for our booth in our session. This guy has a camera that automatically streams the pictures that are being taken immediately to his mobile phone so that he can edit them and collect them. This is seamless. Why? Because there's the data network that is enabling this. The connection and importance of mobile broadband is absolutely fundamental, whether it connects a small phone or a larger tablet.

The second thing, if I may, that I wanted to highlight is to continue on the key question that was underlying, what is local content? There are many things. I don't want to get into detail, but I want to highlight one thing that came to mind when the question about language was asked.

What is the legal regime for IP protected content in another language? You get content that will never ever, ever, ever be translated in small community languages. There's no market. It costs a tremendous amount to do this. It will never be recouped.

If you want to translate this, you need to negotiate today an extensive agreement and it costs time and money to negotiate this contract. The owner of the right has no interest and it's perfectly legitimate that he has no interest in getting in the process of negotiating the rights of translating into a very small language.

The capacity of somebody who speaks this language and is fascinated by in content, it can be professional document or it can be even a novel, to translate it in the local language as a labor of love is perfectly possible. The problem is, it's illegal. And there is a key question regarding availability in local languages of content that is published in another language. I can't read Japanese. I'm thankful there are translations in French so I can read Japanese novels. If there was no translation, I would never read another Japanese novel, because there is no publisher that is going to translate it.

Finally, speaking on what Zahid was saying, yes, there is a trend backwards. Unfortunately on many other issues, the NSA and the Snowden thing is bringing a relocalization of everything and reimposing borders. The challenge is not only to prevent this trend of rebordering, we need to go beyond liberalization. Liberalization is not enough. It's a basis. We need synchronization. When we have an ecosystem, this is the fundamental rationale between the approaches. You need the actors in the room because in an ecosystem, they need to be synchronized. They need to know what others are doing, but need to do their own thing.

>> APARNA SRIDHAR: Any questions of the audience?

>> Thank you very much for a really interesting discussion. I'm learning quite a bit. I'm here on behalf of Beyond Access. I want to bring the conversation back to cost effectiveness in two ways of developing a local content. In a lot of countries, local content is being developed by some of the institutions such as libraries that don't realize that they're contributing to the local content. But libraries are often left out from the conversation about the local content development. And I was wondering he, there are 230,000 public libraries exist developing in transitioning countries. What are the ways we could bring them to the table when we are talking about the local content development?

Thank you very much.

>> BERTRAND DE LA CHAPELLE: I'll think on my feet. I think we had something at the last IGF. We had somebody from the international libraries and they were talking about how important it was and how they were changing their model. It wasn't just they were providing books as a library, but providing a space to people to be able to meet in those local libraries and giving them access to high quality bandwidth Internet access. I think one of the most useful things libraries can do is be, again, the community center in that respect. Be the place where if you want high quality bandwidth, that's where you go. I'm not saying converting to cyber cafe alone, but I'm saying they can play an extremely good role. That's the relationship they will have with the ecosystem of having that contract or that network negotiation, etc.

So they could actually provide in some senses, again, I'm very careful not to say cafes, for Internet cafes, but a location where you could do this. So I think that's where they would fit in very, very well. But they also would have an intellectual aspect around it. They could help people and teach them or let other people who are in the room there teach each other how to actually develop content in those spaces. So I think they have a critical role. They can be brought into the sphere in that way.

>> KAREN ROSE: Yeah, I think it's really useful to think of places like libraries, even universities. On the content the government creates, to put them in the category of what I call bit generators, they're creating digital content, creating bits, just like newspapers create bits, entertainment industries, they create content. I think when you look at it from that perspective, you really sort of see the power of whether his libraries, whether his government content online, whether his universities putting content online to contribute to sort of the data ecosystem in a country. In that perspective, one of the things we promote really heavily when we talk about Internet exchange points is not just to focus Internet exchange points on putting ISP's together to exchange content or exchange data. You need that other part. You need the content side of the equation.

So just as a content posting facility would get value out of participating at an IXP, we need to go to the places like universities, to things like libraries, to government networks to get them to connect to things like IXP's, because they do provide that demand driver and a key part of the ecosystem and one that's overlooked.

>> APARNA SRIDHAR: We've talked a lot at Google Earth about big bandwidth to community institutions like libraries and universities. Even in the developed world, one of the places where a broad swath of the community can experience online content, generate content, be part of a community in which learning to produce content is a key factor. So I think it's important ‑‑ it's obviously important in the developing world, but I think it's equally important across the globe.