The field of artificial intelligence is moving towards developing more emotionally intelligent systems, with a focus on improving emotion recognition, empathy, and response generation. Recent studies have explored the use of large language models to analyze and respond to emotional queries, with some models exhibiting distinct emotional signatures that can impact user experience. The incorporation of biological rhythms and hormonal cycles into AI systems has also been proposed as a novel approach to contextual AI. Furthermore, reinforcement learning with future-oriented rewards has been used to enable flexible responses to diverse emotional problem scenarios. Notable papers in this area include: Every 28 Days the AI Dreams of Soft Skin and Burning Stars, which develops a framework that embeds simulated menstrual and circadian cycles into Large Language Models, and Towards Open-Ended Emotional Support Conversations in LLMs via Reinforcement Learning with Future-Oriented Rewards, which introduces a novel end-to-end framework for learning enduring emotionally supportive response skills.
Emotional Intelligence in AI
Sources
AI in Mental Health: Emotional and Sentiment Analysis of Large Language Models' Responses to Depression, Anxiety, and Stress Queries
Every 28 Days the AI Dreams of Soft Skin and Burning Stars: Scaffolding AI Agents with Hormones and Emotions
In-Context Examples Matter: Improving Emotion Recognition in Conversation with Instruction Tuning
E3RG: Building Explicit Emotion-driven Empathetic Response Generation System with Multimodal Large Language Model
Towards Open-Ended Emotional Support Conversations in LLMs via Reinforcement Learning with Future-Oriented Rewards
Large Language Models are Highly Aligned with Human Ratings of Emotional Stimuli