Hallucination Isn’t a Bug — It’s How Intelligence Works

Author(s): Abduldattijo

Originally published on Towards AI.

Q: Do humans and AI really hallucinate in similar ways?

AI hallucinate. But what if they’re doing it in surprisingly similar ways? In I Think, Therefore I Hallucinate (arXiv preprint, March 2025), researchers compare how errors propagate in human cognition and large language models.

The article explores the intriguing parallels between human and AI hallucinations, suggesting that both may stem from similar cognitive processes. By examining a research paper, the author conducts personal experiments that demonstrate how both humans and AI can fill in gaps of knowledge with confident inaccuracies, emphasizing that hallucinations are a form of predictive overreach rather than random errors. This notion is tied to predictive processing theories in neuroscience, showcasing that such phenomena can emerge under cognitive strain or limited information, and highlighting the implications for both AI development and our understanding of human cognition.

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