The conversation focused on AI-assisted development in WordPress, exploring how large language models like Claude can be both powerful and unpredictable. We talked about the need for guardrails, structured project contracts and reviewer agents, to ensure outputs are reliable, secure, and accessible. We also explored the changing nature of developer work, the increasing playfulness and accessibility of technology, and the challenges of trusting AI-generated code. Several points were raised, including concerns about losing foundational skills, the future of developer roles, and the ongoing evolution of AI within the WordPress ecosystem. Go listen.
