This article is part of the section The Junior Conversationin which specialists from the main universities and research centers answer the questions of curious young people between 12 and 16 years old. You can send your questions to [email protected]
question from aliseda, 16 years old. Center of the Miraculous Virgin (Seville)
Artificial intelligence (AI) is transforming our lives. He is already capable, such as, of detect tumors on mammograms invisible to the human eye or predict death four years in advance. But coudl AI robots one day resemble living beings?
Well, if they aspire to achieve it, among many other characteristics, they should experience
Everyone knows what emotions are, but above all we know how to classify them: joy, fear, sadness, surprise, anger or disgust.
Imagine walking peacefully through a forest. Suddenly, a bear stands in front of you. Within seconds, you’re sweating, shaking, and feeling like your heart is nearly exploding out of your chest. You think about whether the best option is to stay as still as possible or, in contrast, run to a safe place. These “fight or flight” reactions are closely linked to the emotion of fear.
For any emotion to manifest itself, must be satisfied three requirements: interpret or evaluate the situation, produce a bodily response and implement a conduct or behavior.
Let’s go back to the moment when you find yourself facing the bear. Fear invades your mind and, in that moment, the brain and organs work together at full speed to regulate their functioning (it’s called “allostasis”).
This communication is carried out through messaging pathways which are the vagus nerve and the autonomic nervous system. A “subcircuit” in the brain located between the prefrontal cortex (responsible for planning and decision making) and the amygdala (processing emotions and physiological responses) also plays an notable role.
And how do we combine all this information when faced with the emotion of fear? Suppose you decide to stand still in the face of the threat.In the first moments, the prefrontal cortex reduces communication with the amygdala, which in turn tells the heart to slow down. This allows the brain to have more blood flow to evaluate the situation.
After a few moments, you slowly move away from the bear. However, after hearing his growl, you decide to run to safety. At that point, the prefrontal cortex increases communication with the amygdala, which speeds up the heart and delivers blood (along with hormones like adrenaline) to your leg muscles, allowing you to escape.
Although emotions are involuntary, there are situations in which we can “mask” them. Who hasn’t pretended not to be afraid during a class presentation? If you rehearse the content several times and practice your gestures in front of the mirror, no one will notice your fear of public speaking. The presentation was a success and it seemed like you were in control of the situation!
But inside you felt your heart beating at full speed,while your hands were sweating more than normal. Even though you try to hide your emotions, some reactions are impossible to hide.
But why are they so important if we can’t always control them? This involuntary nature has an evolutionary purpose: the survival instinct. Thanks to the emotions and responses that occur between our brain and our body, we are able to survive dangers.
Imagine not getting scared when a bear approaches or not feeling disgusted when you eat something spoiled. And what would our social interactions be without showing joy when interacting with other people?
While human emotions are innate and complex responses, artificial intelligence takes a entirely different approach based on pattern recognition. Identifying these patterns requires gaining “prior experience,” which in your case is a database.
For example, if an AI were to learn to recognize fear, it could incorporate a database with the facial expression of fear or to recognize patterns in heartbeats. This way, if a person had a scared face or a heart pattern associated with that emotion, the AI would be able to recognize it and classify it correctly.
And secondly, you might also learn how to behave in a scary situation (remain still in the presence of a bear).
Without a doubt, AI is capable of learning to identify emotions and how to react to
Can we give robots real emotions?
Currently, machines cannot satisfy all the requirements needed to experience emotions. Even in the recognition and reaction phase they cannot learn on their own. They always need human intervention to be able to acquire “prior experience” (database) on emotions.
and, of course, they lack a living organism that emits physiological responses as we do through brain-organism communication. Although this could be solved by a technology called “neuro-robotics”, which allowed a robot to issue and learn motor responses autonomously with the help of simulation of a small part of our brain: the cerebellum.
Additionally, experts in a branch of science called biotechnology have created “synthetic organs,” which develop much like the organs of any living thing develop. The difference is that they do not interact with the environment, which is important for building on prior experience. If these synthetic organs could be combined with artificial intelligence, who knows if future robots might feel fear, disgust or joy.
In short: Even though AI can recognize and react to emotions, we are still far from creating a machine capable of feeling. Perhaps the most critically important questions that arise here are: should we try to make machines hear? What ethical problems would arise if we achieved this goal? What purpose would it serve?
Emotions are not only automatic responses, but also imply subjective and changing experiences in response to the environment, and also a close relationship with the biology of the living being. So, for now, the day when a robot will be scared if it has a fatal encounter with a bear still seems far away.
The interactive museum Andalusia Science Park collaborates on The Conversation Júnior section.
What are the key challenges in developing AI that truly understands human emotions?
Interview Between Time.news Editor and Dr. Elena Morales, Expert in Emotional AI
Time.news Editor: Welcome,Dr. Morales! Thank you for joining us today to discuss the exciting intersection of artificial intelligence and human emotions. As we dive into your expertise, could you share your thoughts on the recent inquiries from curious young minds about AI’s ability to understand and replicate human emotions?
Dr.Elena Morales: Thank you for having me! It’s a pleasure to be here. Emotions are incredibly complex, and addressing how AI might one day interpret or even replicate them is both captivating and challenging.Many young people, like Aliseda from seville, are curious about whether AI could ultimately resemble living beings in emotional understanding.
Editor: Absolutely. The article highlights that for any emotion to manifest, three requirements must be satisfied: interpreting the situation, producing a bodily response, and implementing behavior. How does AI currently measure up in this regard?
Dr. Morales: That’s an astute observation. Currently, AI can analyze inputs that are indicative of human emotions—like facial recognition software can detect a smile or frown. However, AI lacks the conscious experience of emotions. For example, while AI can predict emotional states based on data, it does not “feel” those emotions as a human does.
Editor: You brought up an engaging point about the physiological aspects of emotions—like the “fight or flight” response when facing a bear, for instance. Can AI simulate such bodily responses in any way?
Dr. Morales: In a way, yes. AI can trigger responses in robotic agents—like increasing the speed of movement to simulate a “running” reaction. However, it’s a programmed response rather than an emotional reaction. Robots don’t have the physical sensations or internal bodily changes that humans experience. They can imitate behaviors but lack the underlying emotional mechanisms that dictate those behaviors in humans.
Editor: So, if AI were to develop more advanced versions capable of recognizing and simulating emotional responses, what would be the key milestones in that advancement?
Dr. Morales: Great question! Firstly, AI needs improved emotional intelligence, meaning not just identifying emotions but understanding context. Then, developing a more sophisticated model of emotional responses—essentially a layered understanding built on the neurological frameworks that govern emotions. Ultimately, creating a platform where AI can engage in genuine emotional communication remains a significant challenge.
Editor: Aliseda also expressed curiosity about masking emotions, something humans frequently enough do. Is there a potential for AI to understand the concept of “masking” emotions?
Dr. Morales: Masking emotions is a conscious choice humans make based on social norms or personal reasons. AI could perhaps recognize when someone is masking an emotion based on inconsistencies in verbal and non-verbal cues. However, grasping the underlying motivations for why someone might mask an emotion is far more complex and still outside the realm of current AI capabilities.
Editor: It sounds like we’re far from AI that truly comprehends emotional nuance. how can young people interested in AI get involved in this field and contribute to its growth?
Dr. Morales: There are numerous avenues for young minds to explore! They can engage in coding and programming through online platforms, participate in robotics clubs, or even take part in workshops on emotional intelligence. Understanding both technology and psychology will be vital in shaping the future of AI. Encouraging interdisciplinary studies will empower the next generation to tackle these intriguing challenges.
Editor: Thank you, Dr. Morales, for sharing your insights! It’s fascinating to consider the future of AI as it relates to human emotions.
Dr. Morales: Thank you for having me! I hope this discussion inspires young individuals to explore and innovate in the realm of AI and emotional understanding.
