Is AI changing how we think? New study raises concerns
Researchers find that relying heavily on AI tools can make writing less personal and more neutral in tone
What happens when humans rely too much on artificial intelligence (AI) to write for them? A new study by researchers from a coalition of West Coast universities suggests that heavy use of AI writing tools may not just change how people write, but also what they say and how they think.
The research evaluation used responses from 100 participants who had to write essays about the relationship between money and happiness.
Researchers discovered that study participants who depended on large language models (LLMs) for their work produced results which showed major differences from the outcomes of those who used no artificial intelligence.
Participants who depended heavily on AI were 69% more likely to give neutral answers compared to others. People who created written work without AI tools showed higher degrees of personal judgement than those who used AI tools.
The research demonstrated that AI-generated content showed more formal style while it lacked emotional depth.
The results which University of Washington computer science professor Natasha Jaques studied demonstrate an underlying issue that extends beyond the current findings.
The LLMs create a force which drives essay writing away from any standard which a human writer would create, according to her observation that she used to describe the effect which makes writing dull.
Heavy AI users produced essays which included personal references because they used pronouns at half the normal rate. The writing became less personal because the writer used general expressions instead of specific personal details.
Many participants also admitted their work felt less creative and less like their own voice, even though their satisfaction with the final output remained similar.
Researchers observed that AI editing tools produce larger editing results than human editors. AI systems perform extensive text rewrites which change original text meaning instead of making minor text corrections.
The discovery establishes fundamental research enquiries regarding artificial intelligence's future impact on human communication and artistic expression. According to Florida State University Computational Linguistics Professor Thomas Juzek, the study demonstrates how AI tools provide more capabilities than basic grammar assessment.
-
Xbox Game Pass Starter joins Meta’s Horizon+ as VR gaming competition heats-up
-
Google unveils cheaper Gemini models as AI competition intensifies
-
Microsoft signs billion-bollar deal to scale Mistral AI in Europe
-
Why China’s internet is obsessed with Moonshot’s viral new Kimi model
-
iPhone 18 Pro September launch: Expected features, A20 Pro chip, camera upgrades
-
France to impose social media ban for under-15s after parliament's approval
-
Google developing 'Frozen v2' chip to hardcode Gemini architecture with 10x efficiency
-
Microsoft adopts AMD Helios: Can it challenge Nvidia?
