#On the front page
The discussion about the scope of technology is common compared to the abilities of the human being.There is talk of the possibility that machines exceed the ability of people and even replace them in some areas.However, there are still elements that computers are not able to fully understand, for example, distinguish emotions and make decisions through them...Although, this is about to change thanks to the advances of the so -called emotional computing.
These types of systems arise from the union of various computational disciplines and cognitive sciences that, at first glance, seem opposite but, when interacting with each other, allow solutions to complex problems.
“It is to take computational science beyond programming.It is about understanding information phenomena to solve problems in a scalable, distributable and replicable way.This allows structure and meaning to the data collected by the cognitive sciences that involve disciplines such as neuroscience and philosophy, which try to understand the human mind, ”said Manuel Aragonés, CEO and founder of Deep_Dive Mexico, a technology signature dedicated to the analysis of the analysisof data, among them, the "identification of emotions".
Although information analysis and problem solving are tasks that artificial intelligence is currently capable of performing, understanding people's emotions and stimuli could help companies determine their strategies.“In terms of business, emotional computation allows you to know if you are doing things right or wrong.You can know if your client is having a good time, as they do in Disney parks where they monitor the emotions of their attendees.It can also be used within classrooms to determine if students are interested in the method applied by the teacher, ”added Aragonés.
Identifying the emotions of those who acquire a product, hire a service or attend an event can mean a great opportunity to create better experiences for users, as happened in an experiment conducted by Deep_Dive Mexico during the Emtech Digital Latam 2019 event where theAssistant faces were captured by a series of cameras installed in the auditorium to recognize the issues that generated different emotions such as surprise, doubt or happiness.
“Instead of putting people to analyze the audience, we use supervised Machine Learning models, a trained neural networks scheme to detect in which part of the video is the face so that it only extracts that part of the image to form a base ofdata adding the place and time when it was captured.In this way I can color the faces according to their emotions;Also know if a person attended the event two or more times, it was part of all conferences and other indicators, ”said Aragonés.
Native not so digital
Add cognitive elements to the different information analysis can also help in customer service tasks, making the interaction between users and companies more efficient to determine the issues that require more attention and immediate action.Such is the case of the Speech Analytics system, which is capable of understanding people's conversations.
"This can work quite well in a call center.We have the case of a company that receives 120 thousand calls per month, and supervising them by personnel can be impossible because it would require complicated working conditions.But this technology allows you to transcribe the calls and take indicators that serve the company to see if the attention is adequate.The classification of these issues within the conversations shows in which there are more silences or even more anger, ”added Aragonés.
Human bias.Despite its scope, this type of computing, for being connected so closely to human behavior, is also exposed to different deviations that detach themselves from the decisions taken by those who have control since, if they wanted to, they could take adirection other than expected.
Also, when talking about applications of neurosciences within computing, it is also necessaryeven more if emotional factors apply that could facilitate the manipulation of messages to favor a few.
An example of the above are the negative practices that have been carried out in recent years in very important social processes such as political campaigns in different countries of the world, where neuropolitics has caused the generation of manipulated information aimed at voters.“This is something that is already practiced, although many times in the campaigns it is said that.Through Big Data and Microsegmentation systems, customary political messages are developed for voters, ”said Cecilia Nicolini, director of Opino Research Center.
Gadgets for return to classes
The expert added that today it is possible to use facial recognition techniques to know what are the emotions that are activated, such as happiness or interest, seeing a political announcement."If this is loaded with violence referring to issues such as migration, it can be understood that this cause.
Another of the methods that are being experienced are the most recent text recognition techniques to generate more realistic bots that not only dedicate themselves to replicating tweets or messages automatically, but can initiate conversations and even incite discussions betweenreal people on some sensitive topic.
“Something that is coming is the new generation bots.With the advance of artificial intelligence and natural language processing it will be possible to have talks with people and even contact voters in a personal way.However, the scale that can be given to that is worrying.The rules of the game have to be established in terms of electoral campaigns but also in the day to day of our democracies, ”Nicolini concluded.
Changes in the Machines and Human Relationship.The natural communication we have daily among people has gradually transferred to the problem -solving models of the different intelligent devices we occupy.However, access to these through complicated interfaces can still mean a problem to achieve the complete inclusion of the population to digital tools, and that is one of the main objectives that, in the short term, can be solved with cognitive computing.
"One of the most important problems is the issue of the interface.The goal is to link us better with the tools to be able to use them.There are some examples that we have already seen as the case of older people who are easier.
Next generation apps
According to the specialist, providing the technology of elements to understand us is very relevant to make it much easier to use it."The objective is that an intelligent assistant can identify levels of despair or urgency, for example, to determine the speed of attention and offer better solutions," he said.
This can cause our perception of this type of technology to change over time and, far from being the surprising factor that is now, it becomes imperceptible to the degree of immersion that can be achieved in the performance of daily tasks and in the solution of problems.
“Our relationship with technology has changed since the invention of the wheel.When it was created, it must have been very relevant but, currently, when you get on a car, the last thing you think is on the wheels, unless they are bursting.Thus, part of the things that seem surprising today will become everyday, ”added Rodríguez.
The ethical aspect.There is a fairly important factor that must be taken into account as longnology progresses and it is the ethics of who has control of it.
While integrating emotional factors into computing has potential for the development of more precise solutions and more natural procedures, it also opens the door to involve human intentions in many of the processes that we carry out today, falling into the field of manipulation.
To the conquest of space!
"I see a lot of potential and challenges.That a computer is able to understand emotions can be something terribly powerful.Technology is a process without moral, but the way we use it does have ethical implications.To achieve a better relationship with this we must ensure that there is an ethical framework around how we are leveraging it, because these new capabilities can be used to determine where you are more vulnerable and, thus, manipulate your decisions and head towards some kind of fraud, for example”, Warned IBM expert.
Add emotional understanding to computational processes can lead to a better understanding between humans and machines and generate more efficient communication and greater benefits that meet our needs.However, like any process where human understanding is involved, there must be pillars that determine the good path that should be given to this technological advance, as Baltazar ends: “Technological change is very accelerated and is increasingly fast.Today we are developing solutions to which we must add bases to prevent someone from using them incorrectly.We have to be aware when we work on this, because it is not only about the progress of technology, but we have to direct ourselves towards good use of it, ”he concluded.
Examples of solutions based on emotional computing
Here are some of the projects that are being developed within the MIT Media LAB, Creation and Research Laboratory of the Massachusetts Institute of Technology, within the APFFECTIVE category.
Bioesnce
It is a device that can be carried as a necklace or on clothes and monitors the emotional state of the person through their breathing and vibrations in their chest, which allows identifying its heart rate.Thus, at the time of witnessing stress levels, the device expels some aromas that help the user to relax.
Live a technological summer
Objectives: to be part of the methods to treat anxiety, stress or sleep disorders.
Applications: reduce symptoms of depression through positive sensations generated through specific aromas, such as citrus fruits.
It works through an app that allows you to adjust how much aromas are expelled, depending on the heart rate.
Bioesnce sigue en desarrollo.Consult more details in: Media.mit.Edu/Projects/Bioessence
AUTOMOTIVE AI AFFECTIVE
This system monitors user activity within a vehicle to determine situations such as humor changes or high levels of distraction so that the trip is safer.
Objectives: Increase travel security by avoiding accidents due to emotions, sleep or distraction.
Applications: the car can determine if you have to take control of yourself to avoid accidents.
Through cameras levels of fatigue or distraction are identified and sound alerts or in the seat are activated so that the person is attentive.
When high stress levels are detected, the user is asked to stop and breathe while reproducing quiet music.The system is already installed in brands of brands like Porsche or BMW.You can consult all the details on the site: Go.Affective.com/Auto
Retro is fashionable...again
Elsa
Elsa (Empathy Learning, Socially-Awareness Agents) is a rather peculiar chatbot because, within its talks, it is not only dedicated to receiving and returning specific or precise information about some data or product but also motivates users to talk about aspects of aspects ofHis daily life with the aim of achieving more empathic conversations between artificial and human intelligence.
Objectives: Support research and mental health treatments.
Applications: Suggest interventions based on patient behavior.
Perform behavioral-cognitive therapies.
Detect individuals at risk of depression or suicide. Elsa (Empathy Learning, Socially-aware Agents), sigue en estado de desarrollo.You can be part of your tests by entering the following link: Elsaneural.net
Executive Commission for Victim Attention
This commission in Mexico has integrated cognitive computing elements to, through text analysis, identify and classify multiple cases files and help crime victims to reach a trial and find those responsible.
OBJECTIVES: Facilitate the review of files and information information to expedite the trial process and the location of those responsible.
APPLICATIONS: Convert documents to PDF images and extract the text thanks to the optical Character Collection that identifies letters and makes them words.
Important names are identified thanks to processes such as Name Identity Recognition to create connection between documents and obtain more precise conclusions.
The tool is now available and has been implemented by its creators: Deep_Dive.
Three trends in cyber risks