Can a chatbot be a co-author? AI helps crack a long-stalled gluon amplitude proof
Phys.org
February 20, 2026
AI-Generated Deep Dive Summary
Theoretical physicist Andrew Strominger was initially skeptical about the potential of AI in scientific research, dismissing early attempts to use ChatGPT as merely clever but lacking substance. However, his perspective shifted when a talented former graduate student, who had left academia for a role at OpenAI, convinced him that physics could benefit from integrating AI tools. This collaboration led to a groundbreaking discovery: the successful use of AI to solve a long-standing problem in theoretical physics, specifically proving a complex formula related to gluon amplitudes, which are fundamental to understanding particle interactions. The work not only demonstrated the potential of AI as a scientific tool but also raised questions about its role in co-authorship and credit in research.
Strominger’s journey from skepticism to acceptance highlights the growing intersection of artificial intelligence and scientific discovery. While he initially believed that physics needed his expertise more than Silicon Valley, he eventually recognized the value of AI in tackling intricate mathematical problems. The former graduate student, now working at OpenAI, played a pivotal role in bridging the gap between cutting-edge AI technology and high-level theoretical physics. By combining human intuition with machine learning algorithms, they were able to crack a problem that had stumped researchers for years.
The implications of this collaboration are profound. If AI can consistently contribute to solving complex scientific challenges, it could revolutionize how research is conducted, potentially accelerating discoveries across disciplines. However, questions remain about the ethical and practical aspects of AI involvement in science, such as how credit should be assigned and whether AI tools should be recognized as co-authors on scientific papers. These issues will likely spark ongoing debates among scientists, ethicists, and policymakers.
This breakthrough also underscores the importance of collaboration between humans and machines in advancing knowledge. While AI lacks the contextual understanding and
Verticals
sciencephysics
Originally published on Phys.org on 2/20/2026