Angeles Inversionistas - Autolang una oportunidad para invertir.

Language is one of the oldest mysteries that remain unresolved. Scientists have been trying for decades unsuccessfully. It's amazing how every child born acquired it in his early and yet we have failed to reproduce this behavior on a computer. Making a computer talk is easy. The challenge is to understand what a human tries to tell. There are a couple of fashonable lines of research among scientists to address this problem. For convenience, we focus on grammatical and statistical techniques. Grammatical technique is based on the concept of grammar. It's what we were taught in school, that tree built on a sentence. This technique suffers to generates more forms than the correct one and that certain phrases have several forms of analysis. This technique assumes that all humans think one way only and same shape. So, even fully understanding the phrase we came not to infer intuitively that tree, so stark and concrete. It's like thinking about the world of ideas of Plato, that was fixed and parallel to our existence. We have preferred to imagine the human as a being capable of acquiring, In statistical techniques, parameters are used to make cells and store odds. As a general rule, We do not usually work with tables, and we are not extra good at maths or odds much less an infant. Nor is it logical to think, that this is an ability unique for language, and not for everyday life. Both techniques have been partially successful. However, none of them reflects the emergence in language acquisition. It's something that every human being naturally gets Regardless of their language or their culture I dreamed for years in face this challenge. I focused my work on it. First solo. And as results came, the team grew. We have been guided by common sense. During these years we've focused on making assumptions and compare them with the computer. Most of the time we crushed, but gradually we selected those techniques that had work. We have progressively distanced from academia. To give you an idea, we conceive of differentiated understanding from language production. We hypothesize that infants first understand the intention and then relates the words with ideas. There are even scientists who defended the impossibility for a kid to learn, for been a seemingly complex process. but they learn, We are the evidence. It was thought that it would take many sentences of training, but to our surprise was not. We have developed techniques that 50 sentences are sufficient for a machine to learn and understand. This technology opens many doors. Too many to address them alone. Surveys, opinion, information retrieval, human-machine dialogues, intelligent search engines, social networks, tracking systems ... and so much more. But because of the crisis, I risk my company and I'm afraid we can't afford much longer to develop this technology. We need investors in the sector of new technologies who wish to embark on a venture capital and give us the final push to make this a reality.