 talking about people who can't speak. I work with people with autism. People with autism have their brains wired a little differently. They've got very poor verbal skills, which means that they find it difficult to pick up language and to use language in communication. On the other hand, they have very strong visual abilities. And in fact, a very famous person with autism once said, I think in pictures, like Google image search. I've been trying to find ways of helping autistic children communicate. In the beginning, when people started working with children with autism, there were picture cards, picture charts. Nowadays, there are apps that are out there. In fact, Avaz, which I developed, is one such app. And these allow children to select pictures to express what they want to say, children with autism. These apps go a long way in giving them the ability to communicate, but it doesn't give them language. Now, those of us who've learned foreign languages or a second language, we know that even if you just have a vocabulary of 50 words, there's a lot that you can say with that. You can take you really far. And that's because that's what language is. It's a way in which you can use a finite number of words. In fact, a small finite number of words to convey an infinite number of meanings. It makes infinite use of finite media. That's what he said. Language is more than just words. It's also a way in which you arrange words to create phrases, to create sentences. And that's what we call grammar. And that's my research. We've already been able to give words to children with autism. And I'm hoping that if I can give them grammar, then we've given them language. Now, this is a very difficult problem. And I'd be the first to admit that I haven't solved it completely. There are several scientific issues here which I have to tackle. First of all, is there this concept called meaning, which is independent of language? Is it something that's tangible? Is it something that we can put down on paper? Second, is that something that can be conveyed with pictures? And even if it can be conveyed with pictures, is there a way that we can translate from that world of pictures, which is the autistic child's world, into English or into other languages, which we are used to talking in? It's a tough problem. And I've solved this problem to a large degree by looking at the way autistic children communicate today. Let's take this example. I told the carpenter I could not pay him. Let's say an autistic child wants to communicate this. I've seen how autistic children are able to communicate some amount of meaning from this sentence. They start with just one word. For example, the word tell. So an autistic child, a high-functioning autistic child, would say tell without any context around it. And then a parent or a teacher would tease out the meaning by asking questions. For example, if a child says tell, the parent will ask, who tells? You tell something? You want me to tell something? Another child will point saying I. And so who tells? I tell. And then the parent will ask the next question. OK, what do you want to tell me? Do you want to tell me something? Do you want to tell somebody else something? Did you tell something else to somebody else? And they'll flip through their picture book, and they'll point to carpenter. At this point, I had this epiphany. I said, well, we've been trying to figure out what meaning is. Maybe this is what meaning is. A bunch of words that you join together in a network of questions and answers. And this was really the brainwave which I had. If you have that as your representation of meaning, and you can then create this engine, which is able to take that picture representation and convert that into English or any other language, then you've got a working picture language. And to my surprise, this has never been done before. Nobody's ever codified grammar in the form of a computer program. And that's what I did. And you can see how it works. For example, we started off with, I tell the carpenter. And so what do you tell the carpenter? You tell the carpenter that you want to pay. Who pays? I pay. And pay whom? Pay the carpenter. And this is what my software does. I can create this picture network. Now I can add annotation. For example, I make a past tense out of the telling. I'm able to add the word can. I'm able to negate that. And voila, this is a statement. I told the carpenter I could not pay him, expressed completely in pictures. Now there are a couple of really interesting things here. First of all, a child with autism or a child with dyslexia, for that matter, has difficulty with linear construction of sentences. You can do something like this in any order you want to. You can start with the carpenter. You don't have to start with tell. And you can build this network up. And the second thing, and this is a cool part, is that it's independent of language. Whatever language the child wants to use as the output, it doesn't affect the way that the child builds the picture network. And if I was able to build a grammar engine for one language, like English, and I was able to build it in Hindi, in Tamil, in Malayalam, in Telugu, in every language of this country, we'd have a language-independent system of communicating. Is that really possible? I mean, the potential is incredible to be able to come up with a way for a computer to understand human grammar. And I ask myself this, is it a problem that can be tackled? I think it is. It's almost a truism now that the ancient Indian mathematicians were good at mathematics. But when you look at the ancient Indian writings, you don't see any mathematical equations there. Why is that? That's because the language Sanskrit is so well codified that the language itself was used as a mathematical language. So I know that it's been done for one language, for Sanskrit. And that gives me a lot of courage that it can be done for every other language in the world. And the second question, which is equally important, if not more so, is it useful? Will it really help? And that's the question that this answers. I saw this article in the newspaper about a year back about a girl called Bhavna, who finished her school. Now she's an entrepreneur. Doesn't sound like a very extraordinary story, but Bhavna is an extraordinary person. All her life, Bhavna's had cerebral palsy. She's not been able to move. She's not been able to talk. Now I worked with Bhavna four years back. That's how I got into Aval. And I told myself, if somebody like Bhavna can achieve all of this with nothing, imagine what people can do with a tool like what I'm building. Thank you very much.