Ask HN: Any success with using LLMs for emotional inference?

I have trouble reading emotions in both myself and others. I've been trying to get better at labeling emotions as I think it could help me manage my own and improve my interpersonal relationships.

Recently I've been using LLMs for reviewing conversations to help me "read between the lines" and infer both others emotions and how they might interpret the emotional content of my communications. I think this has been moderately successful, but it's difficult to tell.

Has anyone else been using LLMs for similar purposes? I'm wondering if it could be integrated into browsers/chat apps in some way, just like automatic translation and transcription.