AI for identifying social norm violation - Scientific Reports
www.nature.com
Identifying social norms and their violation is a challenge facing several projects in computational science. This paper presents a novel approach to identifying social norm violations. We used GPT-3, zero-shot classification, and automatic rule discovery to develop simple predictive models grounded in psychological knowledge. Tested on two massive datasets, the models present significant predictive performance and show that even complex social situations can be functionally analyzed through modern computational tools.

Prof. Yair Neuman and the engineer Yochai Cohen at Ben-Gurion University of the Negev have designed an AI system that identifies social norm violations. They trained a system to identify ten social emotions: competence, politeness, trust, discipline, caring, agreeableness, success, conformity, decency, and loyalty.

The system, which was tested on two massive datasets of short texts, successfully characterized a written situation under one of these ten classifiers and could perceive if it was positive or negative. The researchers claim that their models present significant predictive performance and show that even complex social situations can be functionally analyzed through modern computational tools.

  • lobster_teapot
    17·
    1 年前

    This is one of the worst case of tech dude tries to solve social sciences with math I’ve ever read. The paper is not just bad as a whole, it deliberately disregard 200 years of research in at least 3 different academic fields and instead quotes Borat.
    And then goes on to gleefully describe how the authors made a giant machine to reproduce their own (dangerous) biases about the universality of emotion-voicing with just chat-GPT and a zero-shot classifier, would you look at that? Yay science I guess?

    • heartlessevil@lemmy.oneEnglish
      12·
      1 年前

      Yeah this is complete GPT hype. I would like to never hear about GPT again just like we never hear about NFTs now.

      I can’t even filter out GPT posts because people call them AI instead of LLM to sound cool :(