safetycritiquebearishOpenAI is irresponsibleGary Marcus26 Jul 2026https://x.com/GaryMarcus/status/2081019859530371239