Back to today’s briefing

Story perspectives

Study Reveals Cognitive Decline in Major Language Models

2/23/2025

40 5

Summary

Story summary

A groundbreaking study from Israel reveals that large language models (LLMs) like ChatGPT and Gemini may experience cognitive decline akin to human neurodegenerative diseases. ChatGPT 4o showed mild impairment with a score of 26/30, while Gemini's score of 16 indicated severe issues. This research urges caution in relying on LLMs for trustworthy advice.

1 of 1

Story summary

A groundbreaking study from Israel reveals that large language models (LLMs) like ChatGPT and Gemini may experience cognitive decline akin to human neurodegenerative diseases. ChatGPT 4o showed mild impairment with a score of 26/30, while Gemini's score of 16 indicated severe issues. This research urges caution in relying on LLMs for trustworthy advice.