When AI Research Starts Producing Proofs

OpenAI says a massive agent swarm has produced and formally checked a proposed solution to the Navier–Stokes Millennium Prize Problem, but independent mathematical acceptance is still to come. Jesse examines what that claim could mean for research verification, then covers Meta’s action-taking Muse agent, Google’s evidence of faster AI-assisted cyber campaigns, US allegations of industrial-scale model distillation, Qualcomm and Amazon’s inference-chip partnership, and an urgent N-central security fix. The practical final section looks at more precise editing in ChatGPT Images 2.5 and NVIDIA’s early Rust toolchains for CUDA kernels. Full transcript: https://www.owenonthenet.com/podcast/2026-09-09/ Sources: https://openai.com/index/navier-stokes-solution https://www.claymath.org/millennium-problems https://about.fb.com/news/2026/09/introducing-muse-personal-ai-agent https://apnews.com/article/3a4572eb4cf4e95d8a0dfdad6e6ca065 https://cloud.google.com/blog/topics/threat-intelligence/from-prompting-to-autonomy-the-evolution-of-adversarial-ai https://www.nsa.gov/Press-Room/Press-Releases-Statements/Press-Release-View/Article/4592113/nsa-and-others-warn-china-based-ai-companies-are-distilling-us-frontier-ai-mode https://www.qualcomm.com/news/releases/2026/09/qualcomm-announces-multi-generational-product-collaboration-with https://documentation.n-able.com/N-central/Release_Notes/GA/Content/N-central_2026.3_HF4_Release_Notes.htm https://www.cyber.gc.ca/en/alerts-advisories/n-able-security-advisory-av26-885 https://openai.com/index/introducing-chatgpt-images-2-5 https://developer.nvidia.com/blog/introducing-cuda-rust-two-tracks-for-writing-gpu-kernels