"Artificial intelligence helps us make better decisions.Technology allows processing large amounts of data that the head has no capacity, "said Computer Engineer Carlos Gershenson, participating in the Artificial Intelligence Conference: how to transform information into knowledge, coordinated by members Pablo Rudomin and Claudio Lomnitz and transmittedLive on October 12, through the digital platforms of El Colegio Nacional.
In the session, the investigator of the Institute of Mathematics applied and in systems of theORNAM participated, who spoke about the past, present and future of artificial intelligence;and Saiph Savage, from Khoury College of Computer Science at NortheasternORniversity of Boston, who dictated the presentation artificial intelligence and social networks to create new sources of work.In addition, Ranier Gutiérrez, Cinvestav researcher was present as a guest.
In his presentation, Carlos Gershenson said that three great revolutions can be identified in the history of the human species: the Revolution of Agriculture, Industrial and Information.The first dealt with the domain of matter;the second, of the domain of energy, and the third, of the domain of information.He commented that, in the absence of a definition of intelligence, this concept can be considered as the processing and transformation of information into knowledge with a particular purpose.
He assured that the robots have become prevalent in culture and that the word was first introduced in 1920, in work R.OR. R (Rossum´sORniversal Robots), de Karel Capek.“It was after World War II that the first universal computers were built, which could calculate any type of problems thanks to the mathematician Alan Turing, who said: If we want the machines not to make mistakes, that is not compatible with intelligence.In other words, a requirement of intelligence is to be able to make mistakes and learn from mistakes, but if we hope that machines are perfect that will not allow them to be intelligent."
According to the expert in complex systems for intelligence and artificial life, it was in 1956, at a conference in Dartmouth College, which the term artificial intelligence was coined and lines of research began to propose such as automated reasoning, the theorem test,Games, natural language processing or robots.
He remembered the two winters for which this concept has gone through, the first in the seventies because too many expectations were generated and the machines could not do humans in planned time.The second was in the 90s for similar reasons.
“In the 21st century we lived a third wave of enthusiasm, automated learning arrived, a third version of deep neural networks was reached and many companies made mass investments.For example, China announced that its investment would be 100 billion dollars until 2030 in order to be the world leader.Unlike the previous waves, artificial intelligence is already prevalent in our daily lives, in interfaces, in cars or on websites, everything recommended by Amazon or Netflix is artificial intelligence."
In the words of the specialist, something useful of artificial intelligence is its synthetic method that helps to understand intelligence in general.The method consists of the following steps: the observation of a phenomenon, the construction of a system inspired by that observation, the surveillance of the system behavior and the verification of the theory.“The best way to understand intelligence is to build it, and then we can use it to better understand natural intelligence.To understand how we think of artificial intelligence can be useful."
To the question of whether is good or bad for society? The researcher replied that he will depend on the use, who controls the data that is generated and how open or participatory are the systems.“Now we know that Instagram, Facebook and Twitter have effects on the psychology of their users.If we have capacity as citizens to decide and comment on these organizations, to know how they work, then we can probably decide the positive use of this technology.The more opaque, companies will benefit from their uses."
He pointed out that artificial intelligence helps to make better decisions and the benefits are in what can now be done that does not.However, the risk is that more and more trust this technology.“The question is how healthy will the relationship with artificial intelligence or how sick?Unfortunately, the government does not give it the relevance that it could have, initially it will have to be a more citizen and academic work than government.This is something that should have more relevance in Conacyt and at the federal level."
The expert also referred to the projects he has developed with this technological tool, one of them to detect tax evasion in the Tax Administration Service (SAT), in which he used data of about 7 billion transactions made for four years, and a system inspired by neural networks to generate properties that allowed it to obtain a suspect list. Otra iniciativa fue hacer estudios de velocidades de viento, “necesarios para las energías renovables que el gobierno actual no está apoyando".
He commented that among the perspectives of artificial intelligence it can be assumed that another winter arrives, but unlike the previous ones, in this there are results.“With this technology, like many new ones, there will be jobs that will become obsolete, but that does not mean that people who lose their work cannot find another."
“It's hard to escape this technology, because it is everywhere.We are increasing.By becoming more dependent, in many cases responsibility dissolves."
He added that another involvement is the potential to have a cognitive increase.It is not that machines replace humans, but there is a complementary and not competition symbiosis.“All artificial intelligence that has developed is specific, it means that it is only good to perform a task. Lo que se conoce como inteligencia artificial general todavía sigue siendo un sueño y no está claro que se vaya a poder alcanzar", finalizó.
Are you just starting your Linux journey? Got a friend who's just getting into #Linux? Here's a simple guide I wr… https://t.CO/YUORKLXKAC
— Jason Evangelho | Tech For Everyone Mon Aug 26 16:28:11 +0000 2019
For his part, Collegiate Claudio Lomnitz commented that it is clear that artificial intelligence is here, regardless of what are the anguish in this regard.“The human merged with the technology he has created is not a peculiarity of artificial intelligence or computers, that happened with sticks and axes or any other tool.The human couples to their tools and they extend it and also dominate it.It is a bit ridiculous that people want to insist on a completely separated human vision of technology."
He added that there is an involuntary delivery of data in exchange for an operation that people do not control in the least.“There is a bit of the control of the raw material, but there is no reflection of what the artificial intelligence system for the basic education system implies, for example.If we are already in a world with this tool, this means that teachers have to teach students to use it."
Artificial intelligence and social networks to create new sources of work
Upon taking the floor, researcher Saiph Savage talked about how to create artificial intelligence to build new sources of work.He assured that this technology has created futuristic realities, where there are voice assistants such as Alexa, who can understand almost everything, as well as autonomous cars that begin to understand the world around.
He commented that, although it is true that artificial intelligence has been displacing some workers of his work, he has also been creating millions of new.“These works are concentrated in the fact that employees complete the tasks that are difficult for artificial intelligence.For example, that workers label images because computers and machines do not have the ability to choose them."
De acuerdo con la investigadora del Khoury College of Computer Science at NortheasternORniversity de Boston, otra tarea de obreros que ayudan a la inteligencia artificial es la transcripción de audio y categorización de contenido. “Este tipo de trabajo está al alza, se espera que para el año 2035, el 60 % de la fuerza de trabajo de EstadosORnidos esté laborando con la inteligencia artificial."
He added that this technology has also created a new low social class, because the workers who work closely with it are receiving less from the minimum wage and few growth opportunities.“To be able to create artificial intelligence that produces fair works we have to understand, on the one hand, the different values and needs that exist, and on the other, what is the social problem that you face."
He explained that, with the help of Nussbaum's partial theory of justice, he concentrates on designing artificial intelligence so that workers increase their skills and salaries.“Then, I find a social network of workers who have triumphed, identify them and guide them computationally to share their strategies to be able to develop and achieve better employment opportunities.Then, I share these strategies with other workers so that they can follow them."
The researcher proposes the use of artificial intelligence of reinforced learning, which would allow to know the best strategies of workers to develop their skills and be able to increase their salaries.“The following could be to create a work platform where we put our intelligence and populate it."
Saiph Savage modified artificial intelligence the existing social platforms of large companies to turn them into more just spaces.“What I identified is that, with my tool, the workers were faster and improved the quality of their work.We managed to increase your salary."
“We also made an application where artificial intelligence helped the bosses to be fairer with the workers.That is, where workers' evaluations are considered. Por ejemplo, con un conductor deORBER que haya recibido una mala calificación, porque el usuario estaba molesto con el tráfico, pero el tráfico no es culpa del conductor y le afecta considerablemente esa mala calificación en su ingreso.What he does is evaluate according to the circumstances so that the worker does not lose his income."
According to the specialist, this type of artificial intelligence has also been used to promote creativity at work.“It has worked with different government agencies such as the Ministry of Foreign Relations to offer better services to citizens and their employees, through the creation of intelligent attendees so that government workers can concentrate on more complex and creative tasks."
Savage has also collaborated with the Secretariat of the women of Mexico City, where he designed smart interfaces that allow government employees to better serve women victims of domestic violence.“There is an opportunity to design artificial intelligence that is focused on vulnerable populations in Latin America. La mayoría de las herramientas que existen se concentran en necesidades de primer mundo, pero hay muchos huecos de oportunidades", finalizó.
The Artificial Intelligence Conference: How to transform the information into knowledge, coordinated by the members Pablo Rudomin and Claudio Lomnitz, is available on the YouTube channel of the institution: the collegionationalmxMX.