New Study: AI Models Fail at Basic Human Psychology Despite $50 Million Investment

Author(s): MKWriteshere

Originally published on Towards AI.

Bielefeld University research reveals why AI can’t replace humans in psychological studies

Sarah Schröder knew something was wrong the moment she saw the results.

In this article, researcher Sarah Schröder and her team at Bielefeld University conducted experiments testing the capacity of AI models like GPT-4 and CENTAUR to understand human psychology. They discovered that while humans can discern subtle moral distinctions in scenarios, AI models fail to grasp such complexities. Despite billions invested, AI’s understanding of human morals is alarmingly lacking, equating fundamentally different actions within the same moral framework. This revelation raises critical questions about the reliability of AI in psychological research and has implications across various domains, highlighting the irreplaceable nature of human insight and understanding.

Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI