It is inevitable that we want to face Siri, Cortana or Alexa.We already know that they are robots, specifically, virtual assistants based on artificial intelligence that help us pass what happens.The striking thing is that their voices are attractive and embellish us.We have been made so close that we believe even to know them.
In Uppers we wanted to know who is the father of artificial intelligence, who is the person whose powerful mind has managed to transform our way of communicating with a computer or a mobile and everything that implies.Because if we ask ourselves how our lives have changed the voice attendees, we could ensure that in everything ... and what remains.
This revolution began towards the end of the seventies, when Jürgen Schmidhuber, who was born in Munich and then was 15 years old, dreamed of developing a very smart robot.The truth is that his dream is already quite advanced.Since 1995 he is Idsia scientific director, Institute of Artificial Intelligence of the University of Switzerland, and gives artificial intelligence classes at the University of the Italian Switzerland in Lugano, among other positions.
In 1987 he already described on paper an algorithm capable of improving himself in a thesis on artificial intelligence.Three decades later he has put it into practice with the Long Short Time Memory (LSTM) system, created along with one of his students Sept Hochreiter.This system is used by 3,000 million phones and are part of the products of the world's most leading technological companies such as Google, Apple, Facebook, Amazon or Microsoft.
An imitator of the brain and its neurons
The LSTM mimics the functioning of the brain of the human body whose neurons collect data from images, sounds or words, interpret them and store them as memories.This neuronal network is called Deep Learning or deep learning.The artificial intelligence Schmidhuber system works the same, but instead of neurons there are algorithms.The surprising thing is also that the algorithm is creative.It is based on antagonistic generative networks or gans.They are neuronal networks that are not limited to obeying orders since they interact with each other without human supervision.One network invents problems and the other tries to provide solutions.
When this German scientist published all these ideas towards 1990 it was still soon because the computers were not capable and the databases were scarce to train the algorithms.In 2010 the borders with the Internet were opened and the video games where their gans produce images.These neuronal networks began recognizing traffic signs, Chinese spelling or faces.Then they started talking with people and parking a car.
We still do not know the scope of artificial intelligence.For this computer engineer it transcends biology and even humanity because it will be the robots who are going to conquer space.
What are virtual attendees capable
At the moment we have virtual attendees, a robot that responds to a voice command.As we have explained, as they interact with people they learn and improve their repertoire.They are used both in business solutions and in homes where devices connect to each other through the Internet.Of course, life has changed and facilitated because they are able to ensure security with surveillance cameras, control lights and temperature, notify the expiration of the fridge products, put music and even program events on the agenda and agenda andthe washing-machine.They also recommend restaurants, they dictate the billboard and call your mother if you ask for it.
To date, in the list of the most famous attendees are: Siri, who arrived next to Apple's iPhone 4;Cortana with Microsoft, which forms an inseparable tandem with Windows 10 and is available as an independent app at iOS, Android and Xbox systems;Google Google;Bixby, developed by Samsung and available in the latest versions of its mobiles;or Alexa, the Amazon virtual assistant and the first one who has physical format.