 Hi, everyone. Thank you so much for having me here. My name is Anshal Meena. You can call me An and I'm a product designer and an independent researcher and writer based in the Bay, but originally from LA, so I'm psyched to be back here. I should start my timer. I'm talking about different sorts of divides, but specifically around language divides, and sort of biases around language that exist in our technologies and our technological spaces. I want to take a moment to imagine this next billion group of people who are coming online, and the sort of languages, the sheer diversity of languages that they're speaking. It's hundreds and thousands of different languages, and one language might be Khmer, a Cambodian language. A colleague of mine, a researcher, Ben Valentine, pointed out that when he's based in Cambodia and when he's looking at Khmer websites with certain browsers, the very language, which has its own custom script, appears like this. It looks like boxes. So literally the language of Khmer is invisible to many technologies. It's just one example of how the language that you speak shapes the internet that you have access to, both as a reader and as a listener, as a reader and as a speaker. Take a moment to, when we think about network graphs, when we talk about how the network affects that make up an important part of how social movements and how information is distributed online, there's this assumption in those visualizations that every node in that network is equal. But very often, and you can slice data in many different ways, but the languages that we speak actually limit the networks that we have access to and that we're interacting with. This is a visualization from 2010 by Mike McCandles, who's a researcher who scraped the Twitter data for the languages that people are speaking based on their geolocation and the location that they're tweeting from. And then Eric Fisher then visualized this. And you can see how the languages that people are speaking, each color represents a different language that actually falls along geopolitical lines. And this is not people just sticking to, speaking Italian because they're in Italy, and we're not visualizing what people are speaking based on this map. It's actually, the language itself recreates the map of Europe. And you can expand this with other countries and other regions as well. This can have an effect. One specific example of this, so much so often people talk about the importance of Wikipedia and the importance of open knowledge and open access to knowledge and the ability to contribute to a collective database of knowledge. And Wikipedia has built-in translation features that allows people to contribute language and translation. But again, if you're speaking a minority language, your access to that knowledge can be severely limited. These are the numbers of articles available for different languages. If you're speaking majority languages or languages where people have made a concerted effort to translate that content, give access to millions of articles. It's a great database. But if you're speaking, especially minority Asian and African languages, that number starts to drop significantly. 10,000 for Africans, Tagalog, Kiswahili, down to 100 for even smaller minority languages. We can kind of expect those same patterns, I think, moving back. We can expect similar patterns, I think, with other websites and other sorts of content. And Wikipedia just being one example. And then in addition to reading, it's also the access to voice. I think a lot of us are familiar with the role of the Internet in building social movements and the ability to amplify one's voice. Certainly the umbrella movement in Hong Kong and Black Lives Matter here in the U.S. rely on the ability to broadcast the message, to use hashtags, to amplify a voice and create a pipeline from social media to mainstream media and then hopefully to other audiences. And certainly we've seen, you know, we can think about, like, major hashtags and major movements that have been in English or in a majority language. You know, like a foreign journalist in Kenya, was it a critique of media coverage of East Africa. And then Jesui Charlie is, you know, simple enough French phrase for people to remember and to understand. But there are a number of other movements, and I think the chairs are actually, I hope you guys can see that, but it's kind of fitting, actually. The number of other movements in other languages that are more difficult to understand and more difficult to, and get significantly less attention. Sa Sufi in Congo, there's a Gao movement that's part of the Hong Kong umbrella movement, but is a sort of separate group with sort of different aims and strategies. Lumad Dinaqo, that's in the Philippines, and then this Arabic hashtag means Egypt Delights. It's kind of a parody on a parody hashtag, which I'll talk about a little later. These sorts of movements and conversations are often limited to the language sphere that they're in, because they're often working with minority languages. I just illustrate this even further. I just love this quote from Sarah Kenzior, who's a writer on social justice in Bosa, Middle America and Central Asia. And she's speaking about the kind of quandaries that a Uzbek activist might have to go through to raise awareness for their cause. Uzbek, and I just want to read through the whole description, because it really shows you some of the challenges with amplifying voice when your language is not very well represented in technological platforms, and also there's no pipeline for translating those languages into mainstream and majority media. So the Uzbek activist might know Russian, or Uzbek, and so if she knows Russian, she has to decide whether writing in Russian and potentially reaching an international audience, as well as the 41% of Uzbeks who can read Russian outweighs not being able to reach non-Russian-speaking Uzbeks, or seeming to value a foreign language over one's native tongue. So even the decision to speak Russian over Uzbek, even though there are benefits to that amplification, there are political consequences to not speaking in Uzbek. But then even if she writes in Uzbek, she has to choose which alphabet. And here's where the availability of fonts, typography, and input systems of the Uzbek language have consequences for political action. Does she choose a Cyrillic to reach older generations in Uzbeks in neighboring former Soviet republics who only know the Cyrillic version of their language, or does she choose a Latin alphabet to reach the younger readers who comprise the bulk of Uzbekistan's internet users? So these sorts of dilemmas are much more common when you're speaking a minority language, especially if that language has non-Latin script. So in terms of, you know, as a designer as well and a, you know, product thinker, it's, I'm also thinking about what are potential solutions. And I just, for provocation and for conversation, I wanted to throw out some potential ideas for what, for how we can think about, you know, improving language inclusion, language access across the world. And also here I think in the United States where people are speaking many different languages. I think one of the, one of the possibilities here is crowd sourcing, and crowd sourcing certainly has a lot of problematics. But when you think about the possibilities of translation, machine translation is, can scale very quickly, but it's often inaccurate, right? And anyone who's, who's done translations even between English and Spanish, it leads to much hilarity. At the same time, the translation model as currently exists just simply cannot scale for, for the way that we're, for the sort of content and sort of conversations that need to be translated. And currently, you know, again, crowd sourcing can have its problems. This, this is not a crowd sourced subtitle. This was actually a very famous meme of all your base or belong to us. But it's, it's a sort of risk that happens when fan, when fansubbing communities, you know, translate media popular fansubbing as a fansubtitling. So an example of translating anime movies into English or translating American English movies into Chinese can be done by communities. But you have to have a great deal of faith and trust that those translations will be accurate. At the same time, there, there's, you know, the fansubbing communities can be very, very quite successful. And, and there are more formalized ways of doing crowd sourced translation that are also seemed to be having some uptake. Ted has the open translation project where they, you know, hundreds of, you know, hundreds of volunteers who are translating into hundreds of languages can translate these videos. And that's, and we can see similar examples with, with sites like, like viki.com, where people are can translate content. And I think there is, you know, part of the risk of crowd sourcing, of course, is that the risk of free labor. And I think we need to talk about what, what, what, you know, fair compensation looks like. But at the same time, a broader model for translation can, can be, can help, help ensure that content reaches other languages. EN in, in China is, it's another crowd-sourced site where people are translating, have, were translating articles from English into Chinese as an important way of increasing access for sites like The Economist. It was shut down, and then it's, it's kind of at a, a neutral space right now, but it's, it's an example of the potential for this. And then my own experience is, is building a live platform for translating the Chinese artist Ai Weiwei from Chinese into English with, and his tweets, which, you know, back in 2009, 2010, were, were very, very understood by English-speaking media, despite the fact that he had a major media presence. And, and, you know, this is, is, this model kind of shows that, you know, maybe just five translators can have an impact with 31,000 followers. So it doesn't take a lot, but it does take motivation, it does take interest. And we're trying to productize that with a, at Midan, with a product called Bridge, where, that allows for crowd-sourced translation around social media. And secondly, we need to change the structure. Is it language inequity? It's a, it's a full stack problem. I think you can translate all the things, but if your language is not supported, if your fonts, and this is Amharic, you know, Amharic alphabet with over, I think we'll leave over 300, you know, letters, or alpha syllables. And we have to design better ways to input these languages. We need to design better ways to read them, to access them. And, you know, we, we just need better, a better structure for supporting languages, ensuring that they can be read and input, especially on mobile devices. You know, one, one possibility, and this is a picture of Leo Messi interacting with an app called WeChat, is we also need to think about audio interactions, and audio input. You know, and a researcher friend of mine, Christina, she has pointed out that QR codes are very popular in China as a form of input, because typing in, a very active typing in a Chinese URL can be burdensome, and so it's much easier to take a screenshot of a, of a QR code. So we need to think about different interactions, and especially when we get to languages that may not have a formal written form or any written form, audio, audio interaction and, and oral, oral engagement through technology, I think will be, will be very critical and important. And just, just to close, I'll just give one example of, you know, this has been a, is a full set of color that, that can be exposed through translation and why this can be both a very exciting, but, and, and, and interesting. And, and I think it's, you know, there's really this process of building our global imaginations, that our ability to empathize and interact and, and value people from other cultures in different parts of the world that can often be invisible to the West is, is through, through bringing out that citizen content, bringing out content that can be interesting and valuable. And this is just one, one example of the 2011, the one Joe train, train crash in China was a major train crash where, you know, hundreds of people were killed and injured. And, and this was a sort of event that would have been censored in, in Chinese media, because it was a, it was a sort of, sort of example of a potentially of government mismanagement. But the role of social media in, in, in bringing, in bringing this out was so compelling. And actually telling the specific stories of this is a way of, of highlighting, you know, how, how and why, you know, how and why translation, you know, these, these online engagements can be quite powerful. And, and it really takes translation though to, to highlight these. This was, this was one image where someone photoshopped this into, you know, into this sort of, you know, monster movie. And, and you can see that it's kind of the parody conversations. I'd rather believe this in the official explanation for the train crash. You know, there's different sorts of memes. You cannot escape the blame of profaning the dead. Time to disembark. We're home. There's kind of poetic messages. And this was a friend of mine who had photoshopped a train ticket, starting point hell, destination hell. And that specificity that, of trans, of translating this content and bringing it out becomes important active journalism, I would say. So running out of time. So I'm just going to skip the ones about Egypt. But just to close, I think, you know, as we, we think about the, you know, the, the role of language on the internet. It really biases our, our experience. And, and there's certain, there are a lot of, a lot of risks and challenges there, especially for as, as people from, you know, the global south are coming online. The ability for them to access content for them to contribute to important conversations online will be severely limited. It'll look more like this. And I think, you know, some, some of the most important work we can do in tech is to, to bring it out into languages that they can understand. So thank you so much.