A combination of rapid advances, recursive self-improvement, and agentic swarms are genuinely “spooking people” inside big labs.

---

Link :
https://www.wired.com/story/why-so-many-ai-researchers-think-the-machines-could-kill-everyone/