7 months ago
AI in research boosts individual success but limits exploration
Researchers found that scientists using AI publish more papers, get more citations, and advance faster in their careers.
However, AI also makes scientists focus on the same problems and topics, reducing the variety of scientific exploration.
This means fewer new ideas are being explored, and scientists are interacting less with each other.
The researchers suggest that policies should encourage studying areas with less data and using AI to explore new things rather than just optimizing known solutions.
Researchers using AI publish 3.02 times more papers and receive 4.84 times more citations.
AI adoption reduces the number of scientific topics studied by 4.63 percent and decreases scientist engagement by 22 percent.
Scientists using AI tend to focus on data-rich fields, limiting exploration of other potentially fruitful areas.
AI-augmented research creates 'lonely crowds,' where popular topics attract concentrated attention but with reduced interaction among scientists.
Policy interventions are needed to promote research in data-poor areas and encourage the use of AI for exploration.
- Who
- Researchers from the University of Chicago and Tsinghua University
- What
- AI-augmented research boosts individual success but limits scientific exploration
- Where
- New Delhi
- When
- Analysis published on January 15
- Why
- AI adoption concentrates attention on data-rich domains, reducing exploration and engagement
Key facts
- Researchers
- University of Chicago and Tsinghua University
- Papers Analyzed
- 41.3 million
- Increase in Papers
- 3.02 times more
- Increase in Citations
- 4.84 times more
- Decrease in Topics
- 4.63 percent
- Decrease in Engagement
- 22 percent
- Faster Career Progression
- 1.37 years earlier
Quotes
Authors
Authors of the study
“The team noted that the very models that can generate highly probable outputs are also uniquely situated to recognise the character of surprising data, artifacts, and their scientific consequences.”
livemint.com
“Scientists who engage in AI-augmented research publish 3.02 times more papers, receive 4.84 times more citations and become research project leaders 1.37 years earlier than those who do not.”
livemint.com
Researchers
Researchers from the University of Chicago and Tsinghua University
“This leads to more overlapping research and a contraction in knowledge extent, with scientists converging on the same solutions to known problems rather than generating new ones.”
livemint.com
“Rather than expanding exploration across science, AI concentrates attention on data-rich domains while leaving a growing number of potentially fruitful areas unexplored.”
livemint.com
Researchers from the US' University of Chicago and Tsinghua University in China
Researchers from the University of Chicago and Tsinghua University
“The use of AI could be prompting scientists to 'converge on the same solutions to known problems rather than create new ones', thereby narrowing scientific exploration and engagement.”
livemint.com
