In August 2025, a meetup called “Claude Code Anonymous” took place in London. Organizer Peter Steinberger gathered a group of tech workers who called themselves “Claude addicts” to discuss Anthropic’s revolutionary coding tool, Claude Code. The model, dubbed Opus 4.5, can handle more complex programming tasks, has a larger memory capacity, can run continuously for hours, and manages a set of AI sub‑agents. Anthropic claims that in its notoriously difficult internal test for hiring engineers, Opus 4.5 outperformed every human candidate, sparking widespread debate about how AI will reshape the engineering profession.
By November of the same year, Steinberger released a tool called OpenClaw. It harnesses advances from Claude Code or other programming AIs to let users easily create their own personal AI agents. The tool quickly gained traction. Nvidia CEO Jensen Huang spent over 10 minutes of his keynote introducing OpenClaw and announced that Nvidia had adopted a safer, less error‑prone version called NemoClaw. OpenClaw’s rise marks the rapid transition of AI agents from labs to the mass market, as countless AI companies race to put agents into the hands of anyone with a keyboard or phone, creating unprecedented upheaval in the tech landscape.
