Gergely Orosz · 14 min read · Product Teams
How building software is changing at Anthropic
Read the original on newsletter.pragmaticengineer.com →
AI tooling is reshaping software engineering at Anthropic's AI labs, as teams pursue longer, more complex builds while leveraging faster refactors and AI-assisted verification. In conversations with four engineers, the piece highlights evolving practices and what remains familiar as AI-powered workflows scale across the organization.
Key Takeaways
- AI tooling can dramatically accelerate certain workstreams: Claude Managed Agents took about six months to ship, while Bun's rewrite to Rust went from a year to under two weeks.
- Prototyping is more fluid, but verification and testing are increasingly AI-assisted and can be time-consuming.
- Team dynamics are evolving: design is more ongoing, teams tackle multiple projects, typically limit to two engineers per project, and two-pizza team norms persist alongside continued importance of planning and PRDs.
- Deep domain understanding and cross-functional coordination gain value as AI takes on more hands-on work, reducing fears about job displacement for engineers.
Excerpt
Much-improved AI tooling is changing how we build software, and I want to take a peek into how the future of software engineering may unfold under its influence. What better place for that than with tech’s most “AI-pilled” teams: the AI labs themselves. So, I visited the two leading AI labs to see how teams and engineers do things day to day. In this article and in an upcoming follow-up, I’ll share what I learned about how AI is reshaping software engineering principles many of us are accustomed to – and what’s stayed mostly the same despite the AI wave.…