AI hallucinations are getting worse – and they're here to stay | New Scientist

AI Hallucinations Worsening: A Growing Problem

Recent findings indicate that the latest reasoning models powering chatbots are exhibiting increased inaccuracy due to higher hallucination rates. An AI leaderboard, as per the original article, has revealed this concerning trend.

The Issue: Beyond the Surface

Experts suggest the problem of AI hallucinations is more complex than just a simple increase. This points to a deeper issue regarding the reliability and trustworthiness of these advanced AI systems.

• Key takeaway: New chatbot reasoning models are producing less accurate results.
• Primary Cause: Higher rates of hallucination.
• Expert Opinion: The problem is more extensive than just the increased hallucinations.