Jordi Ribas has been in the United States for almost 30 years. And in 2000 he joined Microsoft, where he currently coordinates the artificial intelligence (AI) area dedicated to product development, both for companies and for the environment of private users. He likes to keep his feet on the ground, and not exaggerate or raise expectations of a technology that he is sure will revolutionize Society. He prefers to be cautious, and acknowledges that getting to today has cost a lot. “ I remember in the United States, when I was studying, there were people who said that in five years we were going to recognize objects. And it took us thirty years ”.
Ribas believes that some of the claims that are made today by some experts and scientists, such as that we are going to have the Singularity around the corner, are very risky. As well as those that say that soon we will have intelligent machines capable of building smarter ones, which in turn will be able to build others with even greater capacities. It is what is called in the world of AI the ‘ machine teaching ‘. “ We are also working and investigating machine teaching, but we believe that it will take a long time to arrive since today it does not work very well. However, it seems like an interesting idea ”.
Despite this, there are still many people who are scared today, precisely because there are those who exaggerate and anticipate what AI can do. “We have to be careful,” says Ribas. “Ethics, privacy, security, transparency are very important. The same face recognition – facial recognition – can be used to do good to humanity, but it can also be used in a negative way. The most important thing, for me, is that artificial intelligence tries to make things easier for people ”.
Also talking about the Singularity, a concept that scientist Ray Kurzweill currently coined at Google, which highlights that non-biological intelligence will surpass human intelligence in 2049 is more than a utopia for the vice president of AI at Microsoft Corporation. He even says that in meetings with colleagues from other large companies with Amazon, Facebook or Google, they think about when each of them believes that the singularity will arrive. ” The one who said the least time indicated that one hundred years, and the one who indicated that more than a thousand years “.
Machines do some things better than humans, but they are some very specific things. General intelligence (GI) is going to take at least ten years to arrive. And there are still today services managed by artificial intelligence that do not work well. “ Speech recognition is still buggy, and natural language processing is very difficult. Digital assistants work very well, but in some segments they fail ”, says Jordi Ribas.
But what most motivates the Microsoft executive, as the head of the development of products that are based on artificial intelligence, is helping people who need it. And he shows me an application that he carries on his smartphone, Seeing AI , a solution for the blind capable of reading texts, describing scenes, identifying objects, expressions or moods.
Seeing AI is a project that was born from Microsoft’s AI for Accessibility , a $ 25 million funded program that aims to take advantage of AI to help more than one billion people with some type of disability in the world. Other projects with the same objective with Soundscape , which analyzes the environment to create a “mental map” of it, describing what happens around and facilitating their steps to the blind with 3D sounds to enrich the context; o Helpico , an application that facilitates communication with children who cannot express themselves verbally through pictograms.
” And it is in all this where I see the disruption of artificial intelligence technologies, and not in Terminator “, concludes Jordi Ribas.
