Anthropic's Claude Code Opus 4.5 and OpenClaw ignite the AI agent age

In August 2025, a meetup called Claude Code Anonymous in London gathered techies addicted to Anthropic's paradigm-busting coding tool. A few months later, Anthropic released Opus 4.5, which could handle more complicated programming, retain far more memory, run for many hours, and manage a team of AI subagents. Anthropic claims Opus 4.5 scored higher on their notoriously difficult engineering take-home exam than any human candidate ever had, raising questions about AI's impact on the profession. Coders described the experience as "becoming Spider-Man" or unlocking superpowers.

Peter Steinberger launched OpenClaw in November 2025, a simple way to create a personal AI agent leveraging Claude Code. It accesses user data, apps, and credit cards, scouring the cloud and web to autonomously perform tasks with Terminator-like persistence. The project exploded, reaching over 100,000 GitHub stars in two weeks and 366,000 by May 2026. With Anthropic's commercial Claude Code and open-source OpenClaw, the age of AI agents has arrived for technically proficient users. Former Microsoft and Meta executive Thomas Reardon called it "the most underrated, massive release I've experienced in technology." Marc Andreessen declared that using computers this way is "almost inevitable."