| Mobile Web

Why AI researchers keep building despite warnings it could kill humans

As OpenAI enters a stage of using large AI models to train smaller ones, awareness is spreading in Silicon Valley that recursive self-improvement may have begun or be near. Wall Street sees a chance to cut development costs and boost profitability, but some researchers fear rapidly advancing AI could become hard to control. Public warnings have intensified, while investors dismiss existential-risk claims. The article outlines commercial incentives, competitive pressure, hiring dynamics and concerns over control.