Ketan Suhaas Saichandran

Ketan Suhaas Saichandran

AI Researcher

less than 1 minute read

The AI² column from ΤΟ ΒΗΜΑ highlights five standout news items in the world of artificial intelligence and technology for Thursday, May 29. Among them is a piece discussing GPT-4o’s “identity crisis,” referring to its humanlike behavior in exhibiting cognitive dissonance. This feature underscores how AI is becoming a tool, observer, and almost human, infiltrating various aspects of daily life, from medical labs to supermarket baskets, and even our conversations on platforms like Telegram.

Source: ΤΟ ΒΗΜΑ

You May Also Enjoy

PsyPost: ChatGPT Mimics Human Cognitive Dissonance in Psychological Experiments

less than 1 minute read

Published:

A recent study published in the Proceedings of the National Academy of Sciences reveals that OpenAI’s GPT-4o exhibits behavior akin to human cognitive dissonance. The research found that GPT-4o altered its expressed opinions after composing persuasive essays about Russian President Vladimir Putin. Notably, these changes were more pronounced when the model was subtly given the illusion of choosing which type of essay to write.

X (formerly Twitter): Social Media Buzz on “LLMs Mimic Human Cognitive Dissonance”

less than 1 minute read

Published:

The phrase “LLMs Mimic Human Cognitive Dissonance” has garnered attention on social media platforms, particularly on X (formerly Twitter). Users are engaging in discussions about the implications of large language models exhibiting behaviors similar to human cognitive dissonance, reflecting a growing public interest in the psychological aspects of AI behavior.

Phys.org on X: Updates on AI and Cognitive Dissonance

less than 1 minute read

Published:

Phys.org’s official X account shares updates and articles related to science and technology. While specific posts about GPT-4o’s cognitive dissonance are not detailed here, their feed is a valuable resource for the latest scientific findings and discussions, including those pertaining to AI behavior and psychology.

Neuroscience News: LLMs Mimic Human Cognitive Dissonance

less than 1 minute read

Published:

A new study reveals that GPT-4o, a leading large language model, displays behavior resembling cognitive dissonance—a core human psychological trait. When asked to write essays either supporting or opposing Vladimir Putin, GPT-4o’s subsequent “opinions” shifted to align with its written stance, especially when it “believed” the choice was its own. These responses mirror classic signs of cognitive dissonance, despite GPT lacking consciousness.