Two teams solved the same quantum crypto problem using GPT-5.6 just three hours apart
Quick Answer
Two teams independently solved the same quantum cryptography problem using OpenAI's GPT-5.6 just three hours apart, raising questions about independent discovery in AI-assisted research.
Quick Take
MIT's Seyoon Ragavan and professors Prabhanjan Ananth and Amit Sahai utilized different approaches to 'unclonable encryption' and are considering merging their findings.
Key Points
- MIT PhD student Seyoon Ragavan and two professors solved the same problem independently.
- Both teams used OpenAI's GPT-5.6 Sol Ultra for their research.
- They focused on 'unclonable encryption' based on quantum properties.
- The case highlights challenges in defining independent discoveries in AI research.
- Reactions to AI's impact on math range from excitement to identity loss.
Source Excerpt
Two research teams independently solved the same open quantum cryptography problem using OpenAI's GPT-5. 6 Sol Ultra, submitting their papers just three hours apart. "If someone mentions an open problem, the first thing is to see if GPT solves it," says one of the researchers. The case raises a question: what does "independent discovery" mean when everyone uses the same models?
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from The Decoder
See more →
An AI model programmed nonstop for 19 days on a single MirrorCode task that cost $2,600 to run
Epoch AI's MirrorCode benchmark reveals Claude Opus 4.7 as the leader with a 56% solve rate, reconstructing a 16,000-line toolkit in 14 hours. Despite this, all models tested struggle with the most complex tasks, highlighting limitations in current AI capabilities. The single task consumed $2,600 over 19 days, raising questions about cost-effectiveness in AI development.
