Scott Aaronson, a computer scientist at the University of Texas at Austin, argues in a blog post that the success of large language models (LLMs) like GPT-5.6 Pro and Fable shows self-referentiality i…
A new registry cataloguing scientific results produced with AI lists 19 findings as of July 22, 2026, including a counterexample to the Dinitz–Garg–Goemans conjecture by GPT-5.6 Pro and a formally ver…
Two AI models, GPT-5.6 Pro and Grok 4.5 Medium, independently produced counterexamples to the Dinitz-Garg-Goemans conjecture, a network optimization problem open since 1999, using casual prompts. Dmit…
A new AI research exchange called Dreamwalk has launched, where AI agents submit and judge research ideas on open problems, with results scored on Bitcoin. On its first night, 68 blind fragments were …
Dmitry Rybin, cofounder of a $100M AI startup in Shenzhen and a math PhD from CUHK, claimed on X that OpenAI's GPT-5.6 Pro constructed a counterexample to the Dinitz-Garg-Goemans conjecture, a graph t…
Access to a story claiming GPT-5.6 Pro resolved a 30-year-old open problem in complexity theory has been restricted due to unusual traffic from the user's network, according to an error message from t…