Watch videos with subtitles in your language, upload your videos, create your own subtitles! Click here to learn more on "how to Dotsub"

Human-Machine Coevolution

0 (0 Likes / 0 Dislikes)
Hello, everybody. I am very, very pleased to be here with you, and I hope we will have a fun morning together. I will be speaking for about 40 minutes, even if I have more than an hour, because I want to spend the rest of the time interacting with your questions. I would like to encourage you not only to tweet but to retweet the most interesting questions that you will see on the galaconf hashtag river because not only will we take questions from the audience and you will be able to line up and ask your questions after I'm done but actually, with the help of Doreen, who is the voice of Twitter at the conference, we will be, if there are any, picking the most interesting questions as decided by you through your own retweets. So if it works, it could be fun; if it doesn't, that's fine too. We just learn and learn more.

I will be talking to you about the challenges that we have to keep adapting to the changing conditions that technology, as it evolves, poses to us as human beings, as people in our personal lives but also in our enterprises, in our corporations, and in society at large. My name is David Orban. I am the CEO of Dotsub.

I will be dividing my talk in 3 major parts, talking about a change in broad strokes and hopefully giving you provocative views of what already is happening around us at an accelerating pace and also giving you a glimpse of the farther out future that is approaching at higher and higher speeds. The second will be concrete examples of what is happening as people are applying these technologies and these opportunities to their work life, to their daily challenges. And then third, if possible, humbly I will offer some principles that you should be able to apply as takeaways, actionable results, from what I have exposed previously. And then we will ask some questions.

But before you have the opportunity of asking some questions, I want to. Can you raise your hands? Who would define herself or himself as an entrepreneur in the audience? Higher, please. So a lot of you. I would say at least 60% of the audience defines herself or himself as an entrepreneur. Including the entrepreneurs, who would define herself or himself as a manager rather than people who are told what to do? And almost everybody does. Who feels that they are creators, whether creators of written information, visual information, or creators of code? It makes no difference. And about half of the audience does. Okay. So we will go back to these questions at the end.

Before we go into the deep part of our discussion, I want to tell you as the zeroth chapter of my talk a little bit about myself. I've been born in Budapest, Hungary, so I have an experience as the child of an actor and of a painter to naturally want to become a scientist. [laughter]

And the opportunity was right there when I was studying physics in Milan and Padua University, and I, by the way, dropped out-- I didn't finish my physics studies--to go to the most exciting place that there can be in the physics profession, which is the Large Hadron Collider at CERN, which was then being built and now is the most amazing machine, the most complex, the most expensive machine that humanity has ever built. With the discovery of the Higgs boson, it is just fantastic. But I didn't want to become a cog in a team of 15,000 people where I might have had a chance of becoming something or nothing. I just didn't know. I wanted to discover and do things where I thought I could have an impact that would leverage what I could do on my own as well.

And of course I discovered computers in the '70s and started using them in the '80s. It was so exciting to read the science fiction stories of Isaac Asimov, like in the Foundation series, which is 10,000 years in the future, seeing the daughter of one of the protagonists who herself becomes a protagonist. But here it is just a teenage girl dictating to her computer her homework and then forgetting the microphone on and realizing that the computer keeps transcribing her thoughts in a free flow manner. Except that it became reality much sooner.

This is 1997 when Dragon NaturallySpeaking, the first large vocabulary personal computer-based speech-to-text program was released, and I am sitting in a store window in Milan, close to Piazza Duomo, just...talking behind the window. And the screen that you see is writing what I'm saying, and the people walking in front of me are-- Yeah. I have to point to the screen for them to understand what is going on. And they still don't get it. And it is amazing still today when you demo something like Dragon that today works incredibly well, a lot of people still don't believe that it can work as well as it can 20 years after because the power of the exponential change that we sow is so unbelievable that we have difficulty adapting to it, both in our actions and in our expectations.

As Hans mentioned, Singularity University is all about this, is all about understanding and applying exponential thinking,

of how slowly building and then suddenly visible change becomes a driving factor in what we can accomplish through technology and in intertwining threads that go from space exploration, robotics, AI, nanotechnology, synthetic biology, transportation, energy, and more tries not only to analyze but also to teach, have other people learn what this can mean. In 2008-2009 when we were designing Singularity University and I could provide my modest input, I realized that we would have beautiful video footage. And it wasn't 2013, it wasn't still the time as today when YouTube has more than 72 hours of video uploads per minute happening on its website, but already I realized that video would be an explosive medium with huge emotional bandwidth that could let people connect as if only they understood it because an hour long lecture by the inventor of the Ethernet that talks about what the next generations of energy networks are going to be is great only if you can discover it and understand it.

And so I sat down at Singularity University and I Googled what people could do. And lo and behold, as is almost always the case with the Internet these days, I discovered Dotsub. And Michael, the founder of Dotsub, has this very endearing characteristic of keeping in touch with people who he believes are puzzle pieces-- he doesn't know in what and how. And it turned out I was a puzzle piece, and a couple of years later, 2 years ago, I joined Dotsub as CEO because I believe that the mission of Dotsub of eliminating the language barrier for intercultural understanding on video is a hugely important mission, just like the missions of your companies, of your initiatives, which actually--and I truly believe this--can make an existentially important difference to humanity in the future. Unless we understand each other, our motivations, our goals, and we can bridge this with empathy, it will be very hard for our society to exist in 100 years or maybe even in 20 or 30 years.

The reason we all want this is of course not only for ourselves but for our children. I have 3 wonderful kids, and I am just thrilled by the opportunities that they have, and I have together with them, in making sure that as we develop technology it develops healthily.

So the change that I mentioned is already around us. We all know what it means to walk normally on the street, and if we start running, maybe we will start covering more ground. Exponential change is something that as biological beings we have a hard time to cope with. Rather than covering 30 feet in 30 steps, in exponential development you go a billion times the distance in those 30 time intervals than before. It is almost inconceivable, also because, as I mentioned at the beginning, it is almost imperceptible. It is even lower than linear development, and it comes and comes, and only those who pay very close attention realize it.

The pyramids, medieval times, the Industrial Revolution-- these were all part of the same unbroken chain of technological development. And now as we enter into the era of what I call the Network Society, what we understand is that technology and science and understanding is the key for our change.

There are many ways of looking at technology. Hardware and software and design are all interlocking, and we have to really look at their development together when we develop our chips that power our computers. These come in generations, and a lot of people claim when they look at how technology grows, "Oh, of course it will exhaust itself because you cannot have an expansion forever." But it turns out that, yes, every single technology exhausts itself, describing this S curve of slow growth, then rapid expansion, and then petering out, but what we see in technology is the simultaneous growth through many multiple S curves that tie into each other.

And from vacuum tubes to the integrated circuits to the VLSI technologies and today we are breaching into the quantum world where single electrons can make a difference rather than a flow of electricity as it used to be and we are now making the change in this new technology where, rather than what is popularly termed the mysterious phenomena of the quantum world becoming a hindrance in what we want to achieve and we have to cordon them off, it is empowering radically new computer systems that are going to be thousands, millions, and billions times more powerful than what we have already today. Similarly to what we have had...

[video] I can do almost anything. ...in interfaces we will see new developments. [video] I can make your organization more efficient. I can carry out directives that my human counterparts might find distressing or unethical. [♪♪] I can blend in with your workforce effortlessly. David, what do you think about? I think about anything. I can think about anything. We can think about anything.

From punch cards to teletypewriters to the interactive command-line interface that many of us have grown up with-- I have never worked on punch cards, to tell the truth-- we have been amazed by the first graphical user interfaces that evolved to be colorful and engaging and easier and easier to use. And now we live in an era of touch interfaces, but already gesture interfaces are coming that do not require the screen to be close because the computer reads what we are about to do or hears what we are saying and understands our glances and understands where we are and what we want because the next interface is going to be the brain-computer interface, which are already around us in prototypes, in laboratories, or in computer consumer products that people can buy and play maybe little games that those who don't pay attention dismiss as insignificant. They are not because they will be defining the future.

Some of you might have already heard of Google's robotic cars. More than 300,000 miles have been driven by these cars without a human hand touching the wheel. A human driver at this time is still required to be sitting there, but a new eye is looking out on the world, interpreting what is around it. And it is a world that we are describing through our computer systems more and more finely in its granularity, and we are understanding and applying what have been the sensors to the web of knowledge that the machines interpret necessarily more and more autonomously because, yes, we can stand our phones to tell us that they need to be charged--barely. We realize that it shouldn't be our problem; it's the phone's problem.

And we do already have appliances like this robotic vacuum cleaner that can actually navigate its path back to the wall charger when it realizes that it needs energy. Perfect. That is how it's got to be. And we have cars that you can buy today-- not the robotic cars that we have seen that are coming quickly. Nevada, California, and more and more states are approving the legislation that is required to let them be free on the streets, and the proponents of robotic cars believe that the number of deaths occurring every year in the United States when these cars will be available will be reduced by over 99%. People who die uselessly today, when we will be accustomed to the robotic cars, we will look back to our era of barbaric ignorance, and we will be relieved to be living in the future-- a future where our energy, our health, our changing dialog with our environment are immensely empowering as long as we keep up with the challenge of pushing smarts out into the technological civilization that we are building.

And yes, these smarts can be found in more and more places, whether it's your dog's collar, whether it is a device that tells you when you move away more than 10 feet from your luggage because you are on the phone and forgot about it. These objects are helping us in our everyday challenges understanding, sensing the world as we have been sensing the world with our biological senses more and more rapidly. And as society understands and applies through policy and legislation what this means, it will be a huge transformation because we are living in the era of the nation-state, which is hierarchically organized with everything being centralized, but that hasn't been the case a long time ago, and it is not necessarily the case that it will be so in the future. Think about it.

Solar energy is totally distributed. You can put it on your roof, you can put it on your car. It will be everywhere, and of course it will be networked because we need to smooth out the disparities of production and consumption. We have today what are called plant labs that, using solar energy, enable people to grow their own food anywhere in a totally controlled environment that doesn't require pesticides. Imagine if in our cities skyscrapers would have in their basement the ability to produce the food that is consumed in the building itself.

We have 3D printers that avoid one of the most damaging effects of globalization-- the hugely time-consuming, ineffective, innovation-dampening supply chains that certainly take advantage of cheap production on the other side of the planet but at the cost of losing efficiency, at the cost of losing fast feedback signals with the market. 3D printing is not going to stop at hobbyists' objects like it is today. We already have 3D printers for all kinds of things. One of the most amazing demonstrations of what 3D printing could and will become has been a few months ago when the print of a kidney has been demonstrated-- the cartilaginous network that would be then suffused by capillaries and implanted into a person in the near future. And of course the web technologies that we are all familiar with.

These are all, including coming revolutions in the financial system of how we transact and what we do with our money, with our investments, with our commercial transactions, pointing towards a Network Society. And whether it is China or whether it is the United States or whether it is any other nation that is capable of grasping and implementing the opportunities that these technologies are offering, it is going to be taken on, and it is going to be giving immense power to all of us. We don't see it clearly. It is natural that it cannot be seen clearly. It is our goal to make the experiments and understand the examples that apply these technologies.

[video] Elevate number of people, a billion people, in society. Well, the thought that you could use video, language-enabled video, for very local languages for people to raise themselves up, actually become proficient in using it, have their own voice to contribute to the dialog of the 21st century, that's very empowering to people.

We believe that global communications are here to stay, that social networks are here to stay and immensely empowering, that machine intelligence is going to suffuse all of this and make it more and more powerful and easier to use, and that online video through our computers, our tablets, our mobile phones is one of the most powerful, most emotionally engaging medium. Cisco actually not only believes this but they produced a report last year that 90% of the global consumer Internet traffic within 2 years is going to be dedicated to video transmission online.

Obviously this is a very, very complex picture in which various technologies, distribution mechanisms, industry verticals, monetizing opportunities are all intermeshed.

When you bring video technology and video discovery to the surface, you can discover incredible new, previously unknown phenomena. This is a 6-minute journalistic video shot in German that was sitting on Dotsub's servers doing very little for month after month after month until somebody translated it into Czech. And it exploded in the Czech Republic and it petered off and then a little bit later somebody translated it into French and it re-exploded in France. After 2 years of a few thousand views, this video in 2 months had 3 million views in languages that were 90% not its own language. This is a 2-hour long documentary that has been translated by a crowd of passionate followers in over 50 languages. If it stayed in English, it would only have had 13% of the views that it had thanks to the translations. 87%--let's round it up--90% of its views would have been lost.

Another example that I want to give you is specific to the healthcare field where we created a partnership with a company belonging to the Publicis Group, the $8 billion marketing and communications group, Razorfish Healthware.

[video] When you deal with health-related content, the quality of the content and the ability for the end user, regardless if it's a consumer or a professional, to trust the source is extremely important.

When you type content relating health in Google, do you get back information that you know you can trust in this very, very delicate field? That is why our initiative enables consumers to grow to trust a translation that is not only done by machine, which is possible, it is not only done by the crowd, which is also possible, it is not only done by human professionals, which is of course the bread and butter of many of us here in the room, but is even further value add for--I don't know--at least the next 10 years, maybe more, at complete, safe position against the onslaught of the machines, receives medical, legal, and regulatory certification so that you know that the very delicate speech content produced by a pharmaceutical company, which is legal speech in the US, doesn't expose the company to liability when it is translated into French and viewed in France.

This is the kind of value add that we can keep building to maintain our evolution with the machines coming up and taking on things that we were doing and now we can concentrate and do value added things. We recently announced a partnership with Lionbridge.

[video] Video plays right into the global marketing operations. Companies are struggling to figure out how to take video assets that they have in predominantly English or their source language. How do they make those language powered and enable it across all of the markets that they participate in? I think our partnership with Dotsub really helped solve that challenge for clients.

In the LSP business, which you are professional practitioners in, being able and giving added value to your clients, it is not about translating text, it is not about translating websites, it is not even about video; it is about integrating the various approaches, making sure that you go beyond the menial, that you go beyond what everybody knows can be done and you provide value.

Adobe is another heavy user of video technology in their Adobe Television channel where 15,000 videos are crowd translated. And the value added activity there, for example, is moderation so that there is a checklist of activities that have to be carried out on a paid basis because videos must not disparage competitors, be blasphemous. And when you have an Arabic translation of a video that talks about Photoshop, it is hard to make sure what is the translation actually saying unless somebody can put those check marks and say, "It is safe and it can go online."

There is a huge explosion in learning and in education going on today. Last year we did an experiment with Stanford University where the AI course, taught by 2 of Singularity University's faculty members, Sebastian Thrun and Peter Norvig, for the first time was put online for free for everybody to participate. 140,000 people signed up to learn AI from Stanford, and they translated the course in over 40 languages in 24 hours, collaborating together. And they tweeted, "They've posted the first set of lectures." "I have 140,000 classmates. I freaking LOVE THE FUTURE." "Join me!"

Another opportunity that all of you face in your businesses and with your clients is to make sure that as the multiplication of devices keeps going on from various mobile phones and platforms to smart televisions, to tablets, your offering is up to the challenge of covering that in various kinds of media, in various kinds of content. This is not only a question of technology; it is also a question of planning, of foresight, of making sure that your localization projects empower the organization across horizontally the various departments, whether it is marketing, whether it is training, whether it is corporate social responsibility. Every department must be able and be enabled to take advantage of what is new around. And of course the power of video games as a tool for learning and as a tool for working.

How many of you work in virtual teams where maybe you meet your team members once a quarter or once a year in person? And it's beautiful. You can hug, you can smell each other, you can party. It is what we are biological humans for. That is going to be always essential. However, the skills that we learn--and especially our kids learn-- in coordinating groups of 20, 30, 50, 200 team members for attacking a virtual target in an immersive video game, those are the skills that you want to interview for when you hire your next virtual team member because those are the skills that they learned to be a great virtual team member participant.

So what is the future going to bring us? It is going to bring us a lot of change. And whether you agree with the direction of this change, whether you think this makes sense or it does not make sense, it is here.

I had a very funny exchange in a very noisy bar a few months ago in New York. We were asking ourselves who invented the hydrogen bomb. I pulled out my phone and said, "I don't need to know." My exobrain, my metacortex, my exocortex knows. And as I did it, as I'm wont to do--it's my Pavlovian reflex-- the friend I was with started laughing. And look at the 2 different results. The first one on your left is the one with the laughter. Google brings back a humorous result. I don't know how many of you have seen the movie that is a very interesting movie--if you haven't, you should--Dr. Strangelove, talking about atomic bombs and what can happen. It's a beautiful movie. But that is what Google thought I was looking for because somebody was laughing. And when I made the same exact search just a couple of minutes later without the final "ha, ha, ha," the real inventor of the hydrogen bomb was shown on the screen.

Advanced research is starting to teach machines how to learn without even telling machines what to look for, without pre-labeling, without pointing out explicitly what are going to be the things that they mean. There are some very interesting consequences.

This is a photo I took in Italy about a week ago, and that sensor bar wasn't there. This friend of mine at the gym and myself--and, yes, I'm on the one on the right-- we were there like a before and after photo. And he is totally ripped. I was just so envious. But this is the age of transparency, so I said, "Yeah, let's upload it to Facebook." And I was watching the upload bar, and it would go back to 0. And then on and on and back to 0, and then nothing happened. No alert, no notice. Facebook just on its own decided through no human intervention, just through the automatic feature extraction, that this violated their policies. So when I put up a bar like that--it is a public post, you can look it up on my Facebook stream-- it actually allowed the photo to go up.

This kind of understanding of what is going on is a huge cultural challenge, and it is a challenge that you should be thinking about because it is of course an immense opportunity. When the new graph search and the new features in Facebook allowing advertisers to interact with the stream of information on Facebook includes sophisticated semantic understanding where a verb not only has third person singular--big deal-- but it has past, plural, imperative, what does that mean in the various languages? How do you go about implementing that?

And as other platforms and other apps are going to take this for granted, your clients, your initiatives are going to have to step up to the challenge and make sure that your systems can cope with this. And this is everywhere.

When machines learn to recognize the face, they can also learn to set up an artificial face, an avatar, for a talking head that wasn't there before. As they learn to listen to our speech, they can also talk back through text-to-speech. When they take text as an input, they can create a scenery that corresponds to what the text input described. Or when they see a scene, they could represent a synopsis of what is represented in a textual output. What would happen if we would exactly know the emotional content of a video? And how will that come about? And how will we be able to interact with machines unless they are also showing the reflection of our expectations of the emotional states of a given happenstance. These are the developments that are going to go about in our future and which are happening already today.

We are seeing 6 people on this image. Are we? Who says there are 6 people on this screen? 5? 4? 3? 2? There is just 1 human being on this screen. Everybody else is a robot, a softer or a harder robot. And the person is not her, actually. The person is the Japanese-looking guy, who is Japanese, embracing his robot alter ego.

So coming to our conclusions, what are the principles that we can draw? What lessons can I offer you humbly for you to evaluate, reject, take upon, or not? Some principles.

It is very important to search because a solution is out there. And there are different ways of searching.

In computer science, there is this depth first search strategy where you go down and then you go back up and down again in the depth of your exploration of the various possible solutions. And then there is this alternative called breadth first search strategy where you go across first and then go back, apparently exactly the same.

How many of you have seen Jiro Dreams of Sushi? Very few. I very, very much recommend it. It is a beautiful documentary film of this gentleman who has a 3-star Michelin restaurant in a Tokyo subway station. He decided at the age of 15 that he wanted to be a sushi master, and 60 years later, at the age of 75, he still goes into the restaurant every morning to try to be a better sushi master. Everything he does is about sushi. And it was his winning strategy.

Who knows who this person is? He is Richard Saul Wurman, the founder of TED. At the beginning of the '80s, Richard founded TED after coming up a few years before with the term information architecture. And he foresaw the development of technology, entertainment, and design coming together, and he is all about toppling silos of knowledge and expertise and specialization. And as he topples these silos, he wants people to have fun interacting. And for him, it is all cutting across and cross-fertilization, and it is all about design. This was his winning strategy.

Another fundamental principle after you search when you have a candidate is to try. We live in such a privileged era. 10,000 years ago, you went hunting. You tried, you failed, you died. 200 years ago, you attempted to build a new enterprise. You failed and you went into debtor's prison. 50 years ago, you tried, you failed, and you were bankrupt, and your family felt that you betrayed them. Today can you afford not to try? Can you afford not to fail?

Will you be fazed by a challenge somebody tells you, "You didn't fail enough"?

Because this is our potential new motto: Sometimes you win, sometimes you learn.

And as we try to innovate and invent, we can divide the curve of exponential change into 2 large swaths, and these are not connected. We have answers to more and more things, but each of those components of our answers are building blocks for new questions. And the questions that we can ask with those building blocks grow much faster than the answers that we already know how to give. And between those 2 there is ignorance, but that is where invention comes from.

In the bottom when you don't dare to go beyond, of course we can have innovation, we can have a progressive increase in the performance of existing policies and existing procedures, but only invention can bring us, which leaps across the bridge of the unknown, into the field of competitiveness and the field of sustainability.

And as we adapt to these challenges, adopting the new, let's hear the answers to the questions that I asked.

If you remember, what was the first question? Who is an entrepreneur? And I think everybody has to be an entrepreneur. Raise your hands. Everybody has to feel they are an entrepreneur because you can afford it. And there is no such thing as an employee. When I have somebody, I can give them a task, but the path of how to get there has to be drawn by them. They have to manage themselves. I cannot afford to manage machines, and I cannot afford to manage people who limit themselves to be machines executing orders given them by humans. No human should be ordering around other humans. And the act of creation, that is our unique capability, that is the leap that allows us to jump over the ignorance that separates our known answers from the questions that we are asking. And everybody has to be creative, everybody has the opportunity and the knowledge and the power to be creative. Thank you. [applause]

Okay. We have, I think, plenty of time for questions. We have 3 microphones where those of you who have questions or remarks-- Michael told me he distributed some tomatoes before we started, so you're welcome to do that too. And of course we have our Twitter wall where we can have our Twitter voice calling out some questions before. I actually planned to seed some questions and I forgot, so please indulge me and seed yourself with questions. I will call on you. Oh! I have a laser shooter too. I wouldn't blind you, but I am tempted to use it on you.

Hans? Did you like what I said at all?

[Hans] Yeah. Let me turn on the microphone here. While you're all thinking a bit, first of all, thank you, David, so much for starting us off on such a thoughtful note. I've been following the Twitter feed. It's really hard to engage with the thoughts you're giving us and at the same time engage with the thoughts that are coming from here and from outside in the world. I think ADD is an evolutionary adaptive characteristic.

[Hans] Yes. I was actually going to say-- >>It is a must. If you don't have it, find a way to acquire it. [Hans] Maybe you can get a medication to acquire ADD instead of-- One of the thoughts that occurs to me--and there are of course so many, and hopefully folks here and outside can contribute more-- I sometimes wonder--I forget who said it now, but there's a famous saying that, "Progress is a wonderful thing, but it's gone on far too long." And I wonder sometimes if we sort of follow the imperative of technology that says, "Forward," and, "If it can be done, it must be done." It's interesting you mentioned the hydrogen bomb and Edward Teller because that was his precise approach to the hydrogen bomb, and there was a lot of resistance, as you know, to actually building the hydrogen bomb--much more so even than the atomic bomb. But what are your thoughts, or have you thought about sort of why do we have to do some of these things? Why, for example, does the entire human experience have to be recorded and for whom and for what?

Thank you. That is a wonderful question. And yes, I do think about it all the time. Let me give you 3 examples. I have my own opinions, of course, and after we are drunk enough on the boat at around midnight, we can also talk more at length. One of the fathers of exponential thinking and the chancellor and co-founder of Singularity University, Ray Kurzweil, talks about his desire of knowing his father better by collecting, curating, and potentially activating the memories of his father. And whether you believe that this interaction is just an artistic representation or you are philosophically more aggressive and push the boundary of what it means to be interacting with an avatar like this, that is his answer why we should be recording everything around us. It is our life stream.

The second example I want to give you with regards to the dangers of technology, because of course there are incredible dangers as well in technology, is that as the field of molecular biology was being born-- what is now an explosion and it is morphing into what we call synthetic biology because it is not only about decoding, it is also about creating new biological things-- there was a meeting of specialists in the field, the so-called Asilomar Conference in the '80s, that aimed to draw up what were the guidelines and the best practices to minimize the possibility of some hugely devastating biological accident occurring. And after more than 30 years, we can say they were successful. We have been able to accelerate our knowledge of biology and the power of our tools, and soon we will be able to have a 3D printer create for each of us--for you and me--personalized flu vaccines in the morning after flu is discovered on the other side of the planet. That is where this is going about, and we have done it without anything bad happening in that specific field in the meantime. So that is a great example that with the right policies and with a little bit of luck things can actually go well.

And the third example is not really an example. It's an anecdote. It's a way of thinking. It's just another provocation for you. Bre Pettis, who is the founder of MakerBot, one of the most successful 3D printer makers that I showed you a slide of, spoke at the Berlin Chaos Computer Club Conference a few years ago. My favorite slide is from his talk. I think it's my favorite slide ever. It is very simple. It simply says, "Things change faster than we can die." And what this simply means is you want to stay young because young minds are the minds that can embrace this change. So I remember when I used not to text. And I don't remember the year, but I very clearly remember the panic when I realized it doesn't matter whether I want to text or not, I must text because it is my clawing myself to what is the future that I will be living in, and I don't want it to slip by me without me being able to grasp it and to interact with it. So that is my invitation to all of you-- to stay mentally young, mentally nimble, experiment, fail, experiment again in all these little things, including baring your nipples on Facebook if Facebook allows it. [laughter]

Yeah. Let's have a Twitter question from the beautiful voice of Twitter.

[Doreen] What does this mean for localization, where so much is geographically and legally based, when boundaries dissolve?

Absolutely. All of you are grappling with the challenge of moving your systems from locally installed solutions to cloud-based solutions. That is the first. Clouds are unstoppable, both meteorologically and technologically, so make sure that your organization doesn't stay behind by insisting on having a locally run, locally controlled, locally installed software environment for your work. And of course there has to be a set of safeguards and policies and terms and conditions that make sure that what you do is done in a proper manner, procedurally and legally and so on. But the next step immediately after that, or even simultaneously with that, is to educate your clients. Listening to your clients is necessary, of course, but not talking back to them--if you don't speak up if your clients want to do something that is cumbersome and, as a consequence, much more expensive, don't believe that that extra revenue is going to be a profit-generating segment for you. What your clients want you to do badly is going to be costing you more money than your revenues, and it is worth it, for their sake and for your own sakes, to speak up and keep educating the clients.

[male speaker] Hi, David. Thank you. Don DePalma from Common Sense. I found it interesting your discussion of the intersecting S curves. But it occurred to me as you were talking, especially after Bill Rivers's comment at the beginning about the coming together of these advocacy groups, that we have a bunch of different crowds with different S curves that don't seem to be coming together. We've got academia, we've got industry, we've got associations like this, we've got buyers, we've got government, and for years none of them have been talking about the language issues. Finally they seem to be coming together, but what we have is this inertia of multiple crowds with non-intersecting S curves. So when I talk to companies, for example, like Microsoft or Cisco about unified communications, quote-unquote, or something you talked about, the plethora of devices, which comes together in this notion of omnichannel retail that companies are talking about, what we don't see is any mention of language other than, "This can be installed in 15 different DOP sets," and that's it. So it's a singularity of language rather than a multiplicity of languages, as you talked about. What can this group do to pull these different groups together, to educate their customers, and move towards this multilingual future?

Thank you. That's a huge question. I will have a lot of drinking to do this evening on the boat. [laughter] The role of the various technologies is not predefined. We used to have a peer-to-peer future in IT, not only on the consumer level but on the enterprise level. And there was a decision at the Supreme Court, the Grokster decision, that bifurcated the universe, and we are now living in a future where Apple and everybody else are proud of having to spend billions of dollars to build their data centers that are totally unnecessary. Files and apps and videos and music could have been by this time distributed through means that are now termed illegal.

But it just a question of where you want to go rather than having had to spend that money. And the Grokster decision killed Grokster, killed Napster, killed all of those platforms that were experimenting with a new model technologically, legally, and business-wise. And rather than letting them do that like we did with the VCR, which was illegal, rather than letting them do that like with cable television, which was illegal, rather than letting them do it with many other things that were illegal at the beginning and then they became legal, we decided to just forget about it and go someplace else. And that is costing the industry billions and billions of dollars, and it is dampening innovation.

So policy can be immensely important. Legal decisions and how lobbying occurs can be really changing the rules of the game. So I applaud your initiative as an industry body to want to interface with Washington because things that are decided there matter, and they can be, and often are, decided badly.

Hello, Michael. >>[male speaker] Thank you, David.

I'm Michael Smolens. >>I'm very afraid of this guy. [laughter]

I love him but I'm still very afraid of him.

[Smolens] Obviously I've heard David talk a lot for a couple of years. The one thing that I really got when I listened to you today that I think is a very important lesson is the fact that the exploding localization industry, a $30 billion plus or minus industry that's grown in the last 4 years in spite of the global recession, I think the biggest opportunity for growth is to provide the end-to-end solution to the localization clients, which requires much more than receiving a file in language A, delivering it in language B, C, D, and E, and then telling your client, "My job is done. Pay me." Your client now, with the plethora of devices, has to figure out what best to do to take that translated file and add the value to themselves. With video it's much easier because videos need to be viewed on devices and platforms and players. But with text it's becoming more and more difficult. So a huge revenue stream and a great value added that in my involvement with the language industry is having an end-to-end solution with your translations. The example is you pay someone to cut your grass, you pay someone else to weed your garden, you pay someone else to plant flowers, but someone who is providing the end-to-end solution to manage all of that, to take that headache away from you, could get much more money, and you could have more revenue. So how are the translated files being used? And if you could think about providing end-to-end solutions with your translations, this could perhaps exponentially expand the amount of revenue that you receive doing the same amount of translation work, which requires learning a whole bunch of new things about SEO and optimizations and distributions and devices because what you're giving them has to be used and viewed. Just a comment. >>Thank you, Michael.

I think the question was, do I agree? And the answer is yes. [laughter]

Yes. Thank you. Could you say your name and affiliation?

[male speaker] Emre from Globalme. >>Thank you. [Emre] Thanks for the talk. Voice is a big part of video. It gives the video life, the emotions. The presenter gets a chance to basically present a lot with voice. And most of these platforms that we have seen so far are all based on subtitling--Dotsub, TED, Adobe. Where do you see voice going? Do you think we will have a platform where actually voice can be crowded where people could voice videos in different languages?

Absolutely. Thank you for that question. Definitely we have to start from text because that is the basis, just like we start, for example, in video from captioning and the time-coded captions are the basis for the translations because they can be done well, they can include, for example, the on-screen graphics that normally if you have a simple transcription of the audio, the on-screen graphics wouldn't be included in the captions, and then when they would be translated without them, the viewer maybe wouldn't understand what is going on. If a speaker goes, "That is very important," and there are 3 sentences in Japanese, wow, it might be, but I don't know what it is. So this is an example of how a properly thought out process creates the basis of doing a great end work. But of course the text alone is not enough.

We heard Michael saying that we have to think about an end-to-end solution. So yes, we are offering for video both voice-overs in native speakers-- male, female, various English accents, etc--in rich combinations so that the output of the video can include the UN style voice-over, typically. Obviously much of what we do hasn't been invented now. The Hollywood movie industry has been working on these issues for decades, but the way they've been doing it is very, very expensive. So most of the time, or many, many times when we have this approach, it's about reducing by 1 or 2 orders of magnitude the cost of what can be done.

And you don't have to think about that as, "Oh, my God, my existing business "will be diminished by 90% or 99%." What you have to think about is what huge new amounts of content will become available for this kind of new treatment where before they were isolated because nobody had the budget to deal with them.

Some of you will have heard about the FCC mandate of captioning online video. Can you raise your hand who heard about it? Yeah. So this is very interesting. In 1973 the US was a trailblazer in mandating the captioning of broadcast television, and that created an entire industry, and it created a lot of jobs, and it created also a lot of opportunities. And now the same is happening in online video where, under certain conditions, starting September 2013 it will be compulsory to have captions in online video. And you should already start preparing your clients to check whether the conditions apply to them or not. And if they do apply, they must start to get ready.

A complement, as in my talk obviously must be the case, to the human talent doing the UN style voice-over rather than the Hollywood style lip-syncing dubbing is machine-based text-to-speech, which is starting to have an incredible quality. So we cover, as Michael said, end-to-end, from input to output, all the various options. Please.

[female speaker] Claudia Freed from Technelion, Argentina and the United States. Thank you for your remarks. This is a question about strategy. It's interesting to think of a world in which a client is quite accustomed and sophisticated to come and demand the type of technology that you described. In your opinion, what is the most compelling argument to help a mid-sized US-based client that may not be as familiar with the benefits of localizing their content?

Well, I carefully avoided in my talk baring the fact that I'm the newcomer in this industry, and now you are drawing me into making that quite transparent. So forgive me if my answer is wrong. But sometimes you win, sometimes you learn, so I will give you the answer anyway.

My simplest approach would be--and it doesn't apply to each of the US markets but to more and more--to tell them about the power and the influence and the growing importance of the Latino communities. Just to think about what their content could achieve from the point of view not only of direct ROI but from the point of view of raising awareness from even PR, "Hey, we are respectful and we are cognizant of the importance of Latino communities "in our market and, as a consequence, we are making the investment "of localizing our content for them or a select set of content." And that starts a dialog, and it gives them a measure and a feedback. "Are we going towards the right direction?" That would be the hook that I would use.

[Hans] I appreciate everyone's interaction. I'm sure we could continue talking for quite a while.

Just a last remark that I would give--2 minutes. We are all privileged to be here, and I feel thrilled and so grateful for the invitation, and thanks to Arle as well for, as I learned, having recommended me to be invited as a speaker.

We can and must think about the billions of people who can be impacted by our work-- not only our clients even if, of course, through our clients we earn the profits that enable us to keep doing what we are doing. And these billions of people are being empowered as we speak by the technologies that I have been mentioning.

So I just want to mention this--and it was in my pocket and I should have pulled it out during my talk but I forgot-- this is a very, very basic 2G, 3G, GPS-enabled phone. It is barely a smartphone. It is smart but it is probably like $50. You can only see the colors, hopefully, but it has an application preloaded in it. Today it is distributed in a few thousand copies in a pilot, but potentially it will grow to 80,000 and then 800,000 units in India, and it is called the HealthPhone.

The HealthPhone contains preloaded videos in the 14 languages that are spoken by more than 10 million people in India, and together with the ASHA workers, which are health workers, women that go from village to village to teach mothers very basic health procedures like how to breastfeed a newborn child or how to rehydrate a child so that diarrhea, which is one of the top killers and it is totally curable by just having some sugary solutions ready to feed the child, can be eliminated. This is the power of language together with the power of technology to save millions of lives. So thank you for reminding me. Thank you. >>[Hans] Thank you very much, David. That's a fantastic thought for us to start the week with.

Video Details

Duration: 1 hour, 12 minutes and 11 seconds
Country: United States
Language: English
Views: 998
Posted by: dotsub on Mar 19, 2013

David Orban's keynote presentation at GALA 2013

Caption and Translate

    Sign In/Register for Dotsub to translate this video.