Robots and AI - The Robotic Future

Goal 2020: this will be the robots of tomorrow 

They dare to serve the public in hotels, monitor and guard shopping centers and lend a hand - or rather, a pair of precise forceps - to surgeons in certain operations. More and more are the spaces in which we discover a robot and a human interacting with each other and it seems that in the near future it will not be uncommon to find a machine working side by side with a scientist or as a helper accompanying a restaurant chef. 

This is at least predicted by the American consultancy Merrill Lynch , which estimates that by 2020 the robotics and artificial intelligence sector will reach a value of 153,000 million dollars (146,000 million euros). Gartner analysts, for their part, point out that by that same year robotics will be among the main concerns of the CIO of every company.

But what will the robots of tomorrow look like? What relationship will they have with humans? Will they become our co-workers and friends or will they be smart enough to fear them? 

The future today:

A couple of months ago Baker Hostetler, an American law firm, signed law, for one of its teams ROSS , a robot with knowledge of bankruptcy that helps lawyers with investigative tasks. ROSS reads the various regulations, gathers evidence, and even dares to make inferences and provide evidence-based answers.

For this, ROSS uses artificial intelligence algorithms Watson's , the computer system created by IBM capable of understanding human language and interacting with them. Algorithms that also allow you to learn from the attorneys who employ you . But despite being endowed with artificial intelligence, ROSS is not a robot with a physical presence, rather a program.

More body presents Da Vinci , a robot made up of four arms that helps surgeons to perform operations such as operations for prostate cancer, a gastrectomy or endometriosis. Since it was launched in 2000, Da Vinci has participated in 3 million operations . In Spain it is already present in 24 hospitals, 13 of them public , and this thanks to the benefits it entails for the patient: a faster recovery and less blood loss during the intervention. 

Both ROSS and Da Vinci are part of the technology that a company has. Artificial intelligences that, according to Rob Nail, founder of Singularity University , have been made possible by the ever-shrinking size of sensors, the greater processing power of computers, the widespread use of GPS technology, and the advancement in programming languages, which has lowered the price of owning a robot.

As noted in a report by the US asset management firm Piper Jaffray , if in the past owning an industrial robot cost a company between $ 100,000 and $ 500,000 (between 95,000 and 479,000 euros), currently the price of an intelligent machine is between the 20,000 and 50,000 dollars (19,000 and 47,000 euros) , a cost similar to having a vehicle.

However, today there are also robotic solutions for consumers. Of course, not all have the appearance of robots. As Joan Oliver, director of the explains to JUGUETRÓNICA Institute of Robotics for Dependency , , an example would be Amazon's assistant, Echo .

This, in particular, "can be a help for dependent people, it warns you about the weather or to be careful with the clothes you wear, and listens and answers questions," he says. To which he adds that for a little more money we can get a machine with a face capable of interacting more with the user "and without the need for great artificial intelligence ."

The Robotics Institute for Dependency is part of the Ave María Foundation, an association in charge of caring for people with intellectual disabilities, either at home or at the foundation's residence.

Among its projects in development, the Never Alone project stands out, a computer program that allows managing communication between the different organizations and people that surround the dependent person, in addition to reminding them of certain activities to carry out when they are at home, such as taking out the garbage. after a light came on next to the trash can. "Simple systems, low cost, little or no intrusive and specific for each person", he explains. 

Despite the fact that the solution of the Institute of Robotics for Dependence is to make the home of the dependent person intelligent in order to help him, Oliver is clear that in the future this system will coexist with the presence of a robot .

"When you have to interact with the person, such as doing an activity, a game or trying to socialize, a robot is an embodiment of this software and an element to interact more effectively," he says. To which is added the fact that caretaker robots have sensors that help detect the user's emotions. 

Emotions in a robot:

Because in order for a robot and a dependent person to interact, it is necessary that the latter understand and know how to recreate the emotions of the human being . And this to get the latter to like the company of the robot. This is confirmed by Helena Matute , director of the Laboratory of Experimental Psychology at the University of Deusto.

"For humans to like to interact with the robot, it has to have a fluid and agile conversation and the more it understands us and the more empathetic it is, the better," he explains. And he gives as an example the moments when a person feels distressed. "If he is sad and the robot realizes that he is sad and asks him what is wrong, he will feel better with him than with a cold machine."

Something in which also agrees Ricardo Sanz , doctor in robotics and artificial intelligence from the Polytechnic University of Madrid, . "People are used to interacting with people and if we have a humanoid robot that is capable of putting appropriate faces, the better," he says.

For some time now, the European Commission has invested, together with European companies and academic entities, in various projects on robot carers for the elderly or disabled. In all of them, in addition to creating the robot, the researchers in charge of a certain prototype study the relationship it maintains with an elderly person by installing the machine in their home.

"It is being seen that when the robot is taken back to the university, people have a hard time  , as if it were a pet that they have grown fond of," Matute tells us. Hence, the optimal level of empathy that a robot should have is also studied. 

 

But the fact that a robot is capable of detecting emotions and reproducing them does not mean that the machine understands them and that is something that, according to Sanz, actors know how to do. "An actor who cries in the theater does not have to be sad and a robot who makes a face or a gesture that we associate with sadness does not mean that the robot is sad."

Something that, however, can be achieved in the distant future. According to Sanz, research in robotic emotion is beginning to take its first steps , although we must first understand how a human being feels emotion. "If it is not known to humans, we can hardly build machines."

Risks of Artificial Intelligence:

Despite the fact that many consulting firms estimate the boom that the robotics sector will experience in 2020, Gartner analysts assure that it will never be possible to create a completely autonomous robot . That will neither be possible nor desirable.

The consultant recommends that the machines of the near future are capable of learning the predictable, that they can be directed with precision and in a controlled way and that the human being be the last person responsible for the consequences of their use.

Because the fact that an artificial intelligence is capable of learning by itself - as is the case today - carries risks. Among them, for example, that the machine knows how to act and manipulate following the expression of our emotions. According to Matute, the robot learns by following the stimulus-response process, so it can learn to react in a certain way to one or another emotion and act according to the objective it pursues.

Without forgetting that the same ability to learn that today's machines have makes it possible for these machines to surpass us in intelligence in a short time . Hence, the psychologist deems it necessary for public administrations to think of measures to avoid such risks, as has happened in the United States and the United Kingdom, “where they are going to create a commission to continuously observe how artificial intelligence algorithms are developing ”.