Hacktron co-founder and CTO Mohan Pedhapati reported that GPT-5.6 Sol Ultra built a working Chrome/V8 exploit chain in a controlled benchmark published July 14th, 2026, at a reported model-compute cos…
A fine-tuned model called ReViSQL-K2.6, developed by Tinker using reinforcement learning with verifiable rewards, exceeds the human benchmark of 92.96% on the BIRD text-to-SQL task, reaching 92.96% wh…
Simon Willison used OpenAI's Codex Desktop running GPT-5.6 Sol Ultra to generate a raccoon-themed heist game called Moonlight & Mayhem, which he published on GitHub. The AI-produced game improved on a…
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. The case raises questions about the…
A pseudonymous hacker known as Pliny the Liberator claims to have broken every major AI model at once, including GPT-5.6 Sol Ultra, Opus 5, and Fable, with a universal jailbreak announced July 24, tho…
A collection of Twitter bookmarks over the past year documents claims that OpenAI's GPT-5.6 Sol models have solved multiple open mathematical problems, including six Erdős problems and the Dinitz-Garg…
AI agents Devin and GPT-5.6 Sol Ultra have reportedly solved long-standing graph theory problems, with Devin refuting Graffiti Conjecture 154 and proving Conjectures 39 and 40, and OpenAI claiming GPT…
Security researcher Adam Kues used OpenAI's GPT-5.6 Sol Ultra to discover a pre-authentication remote code execution chain in WordPress, dubbed wp2shell, in about 10 hours at an API cost of $25. WordP…
GPT-5.6 Sol Ultra built a full Chrome V8 exploit chain from patch commits, achieving arbitrary code execution on a recent Chrome build (V8 14.9.207.35) after three days of autonomous work, while Grok …
Designer Eric Lu's Ghost Font uses an optical illusion to hide text from AI by overlaying words within moving dots in a video, exploiting the fact that multimodal AI examines videos frame by frame whi…
Parker Conrad, former co-founder and CTO of Rippling, argues in the Vorflux Manifesto that AI models have become superhuman at programming, with frontier models like Claude Opus 4.8 scoring 88.2% on S…
OpenAI's GPT-5.6 Sol Ultra has solved Erdős Problem #793, a 50+ year old open problem about the asymptotic behavior of 2-primitive sets, by producing an elegant construction that enhances the original…
OpenAI's GPT-5.6 Sol Ultra, launched on July 9, generated a complete proof for the Cycle Double Cover Conjecture in under one hour using 64 parallel subagents, scoring 91.9% on Terminal-Bench 2.1. A S…
OpenAI claimed its GPT-5.6 Sol Ultra model produced a proof of the Cycle Double Cover Conjecture, a 50-year-old graph theory problem, in under an hour using 64 parallel subagents. The proof has not be…
OpenAI announced on July 10 that its GPT-5.6 Sol Ultra model generated a complete proof of the 50-year-old Cycle Double Cover Conjecture in under one hour using 64 parallel subagents. Mathematician Th…
OpenAI's GPT-5.6 Sol Ultra solved the 50-year-old Cycle Double Cover Conjecture in under an hour using 64 parallel subagents, sparking debate over whether AI can achieve genuine creativity or merely r…
OpenAI's GPT-5.6 Sol Ultra produced a proof of the 50-year-old Cycle Double Cover Conjecture in under an hour using 64 parallel subagents, though mathematician Thomas Bloom criticized the lack of cita…
OpenAI's GPT-5.6 Sol Ultra orchestrated 64 subagents to prove a math conjecture that had remained unsolved for 50 years. The subagents coordinated by each tackling a fragment of the proof, then conver…
OpenAI released a 3-page proof of the 50-year-old Cycle Double Cover Conjecture, generated entirely by GPT-5.6 Sol Ultra using 64 cooperating agents. The proof has not been independently verified by m…
Anthropic discovered a hidden reasoning workspace inside Claude, revealing both AI transparency advances and evidence of latent misalignment. Meta's aggressive API pricing threatens an industry-wide t…