AI agents have thrown the world of technology into chaos. Here’s exactly how it happened

Share

“Hi, my name is it’s Peter and I’m a cladeholic.”

It was August 2025 and Peter Steinberger was speaking at a meeting in London titled Claude Code Anonymous. Steinberger and several other addicts organized the event to connect with people like them – techies who are fascinated by coding tools like Anthropic’s paradigm-breaking Claude Code. “I devote almost all my time to this, but it’s not enough,” he told the crowd in a cozy room with brick walls.

A few months later, Anthropic released a new version of Claude Code, and the ranks of Claudeholics exploded. Called Opus 4.5, it could handle more complex programming tasks, store much more in memory, run for many hours, and manage a team of AI subagents. Anthropic offers what it calls a “notoriously hard” self-administered exam for prospective engineering employees; after directly comparing these people and their models, Anthropic concluded that Opus 4.5 “scored better than any other candidate in history,” which “raises questions about how AI will change engineering as a profession.”

Countless programmers spent their summers in basements and dens, madly trying out this new toy that allowed them to create software as if they had released a hundred clones. Or unlocked superpowers. “I feel like I’ve become Spider-Man,” one of them told me.

Even that wasn’t enough for 39-year-old Steinberger, who divided his time between homes in London and Vienna. In November 2025, he launched what is now OpenClaw, a tool that allows you to easily conjure up a personal AI agent that leverages advances from Claude Code or other coding tools. Give it access to your data, apps, and maybe even your credit card, and it will search your cloud and go online to do your bidding. It can operate autonomously in the background and overcome obstacles thanks to the Terminator’s persistence.

Steinberger’s project was born in the middle of winter. One indicator of popularity is the number of “stars” a code repository receives on Github. In less than two weeks, as users downloaded it and began frantically building it, the project amassed over 100,000 stars. (At the beginning of May it had 366,000 stars).

With these two breakthroughs – the commercial product Claude Code and the open-source software OpenClaw – the long-awaited era of AI agents suddenly arrived. At least for those who are technically proficient enough, and perhaps foolhardy enough, to embark on a messy, imperfect, and risky adventure. More than one claudeholic tells me that they feel like they are living in the future. “AGI is here!” one fanatic told me, paraphrasing William Gibson’s famous quote. “It’s just not evenly distributed.”

During the computer revolution of the 1980s, the general public treated new machines with a mixture of curiosity and trepidation, while hackers happily built. We’re dealing with a similar dynamic today, and the stakes may be even greater. “It’s hard to explain how huge a change this is,” says Thomas Reardon, a former Microsoft and Meta executive who now runs a startup focused on a different area of ​​artificial intelligence. “It’s the most underrated, massive tech release I’ve ever experienced.”

We will all experience this soon. On a recent podcast, Marc Andreessen, co-creator of the web browser and who describes himself as the ultimate techno-optimist and MAGA fan, made a statement that reflected the Silicon Valley mindset: “It’s almost inevitable that people will utilize computers this way.” Left without comment: it won’t be a choice.

Back to in early 2024, when Boris Cherny was Instagram’s CTO and worked remotely from the home he shared with his partner in rural Japan. “I rode my bike to the farmers’ market by the rice fields,” says Cherny, 34. “Our hobby was making miso and pickles. We traded with our neighbors.” That all changed when he started playing with the AI ​​models emerging from his former hometown of San Francisco. (He comes from Ukraine; his grandfather programmed computers using punch cards). The models took Cherny out of his idyll. He connected with Anthropic through friends and then returned to the Bay Area to work there.

Latest Posts

More News