Can AI truly understand emotions and ethics? This article examines the limitations of affective computing and empathy algorithms, and raises philosophical and neuroscientific questions about whether AI can replicate human morality and emotional systems.
Artificial Intelligence Encyclopedia.
You can find the artificial intelligence knowledge you need at aiseries.oopy.io.

Lisa Feldman Barrett's Theory of Constructed Emotion proposes a fundamentally different perspective from traditional emotion theories.
Barrett argues that emotions are not predetermined biological categories, but rather phenomena that the brain actively constructs by synthesizing past experiences, cultural concepts, and current sensory information.
The brain constantly predicts and interprets the body's changing internal states and sensory input from the external world, giving appropriate emotional meaning to the situation.
In this process, language and social learning play a crucial role in helping individuals form emotional concepts and categorize their sensory experiences into specific emotions.
According to his theory, one of the brain's primary functions is to coordinate and regulate the body's systems, and this is accomplished through predictive coding mechanisms.
The brain predicts the current situation based on past experiences and updates its internal model to minimize the discrepancy between the prediction and actual sensory input, that is, the prediction error.
--- From "01_Brain Neuroscience Principles of Human Emotions"

Hume presented a fundamental challenge to the rational-centered ethics that was mainstream at the time, and his arguments later had a great influence on various fields, including moral psychology.
Recently, the role of emotions in moral judgment processes has been emphasized in contrast to the traditional rational approach.
Joshua Greene and his colleagues proposed a "dual-process theory" of moral judgment and published research supporting it.
According to Green's dual-process theory, deontological moral judgments are primarily driven by automatic, emotional responses, whereas utilitarian moral judgments are driven by controlled cognitive processes.
According to his research, certain types of moral dilemmas, namely personality dilemmas, tend to elicit salient and automatic emotional responses, which are quickly linked to specific judgments, such as the inappropriateness of an action.
--- From "03_Moral Judgment and Emotions and Feelings"

In a series of neuroimaging studies, Green and his colleagues investigated whether differences in brain activity and reaction times associated with moral judgments existed in two versions of the well-known trolley dilemma.
In the first case, the trolley dilemma, or the switch case, most participants answered positively to the question of whether it was morally permissible to change the direction of the trolley when it was running towards five people and sacrifice one person instead of five.
The second case was the footbridge dilemma, in which participants were asked to judge whether it was morally permissible to save five people by pushing one person off a bridge into the path of a tram.
In contrast to the trolley dilemma, most participants gave a negative answer.
Why do these differences arise? Green and his colleagues hypothesize that these differences in moral judgments stem from different psychological and neurological processing pathways in the brain.
--- From "06_Empathetic AI"

As AI assumes a role equivalent to that of a de facto decision-maker in various societal domains, the moral legitimacy and accountability of AI's judgments are emerging as important public issues.
Especially in the fields of medicine, education, and transportation, which are directly related to life, safety, and rights, clear principles and guidelines are required regarding the ethical standards by which AI should make decisions.
For example, when a medical diagnostic algorithm analyzes patient data to suggest treatment options, transparency and fairness are essential in determining which ethical values—patient respect, equity, and privacy—were prioritized, along with the model's training data composition.

--- From "09_Ethical Standards and Social Responsibility for Moral AI"