The broader trend is that AI coding tools blur traditional roles, enabling non-technical workers like product managers to contribute directly to product development, a shift echoed by leaders at Figma.

AI is blurring traditional tech roles across the industry, with companies like Figma and LinkedIn adapting practices to the change by encouraging experimentation and restructuring roles.

AI coding is redefining job titles and responsibilities, with Figma’s Dylan Field noting non-engineers are dipping into building; LinkedIn has begun training hires as multi-disciplinary builders through an associate product builder track.

PMs can prototype smaller UI features and hand them off to engineers for final review, while complex infrastructure work and large, high-risk projects should remain in engineers’ hands, suggesting roles may become more fluid over time.

A PM describes shifting workflow by building smaller UI features themselves and then handing code to developers for final review, illustrating a collaborative learning approach with engineering teams.

AI tools enable PMs to move from coordination to product ownership with execution capabilities, though core infrastructure work remains the domain of engineers.

Arnovitz’s experience highlights a shift toward hands-on building and prototyping in product management, though complex systems still require engineers.

Experts suggest the trend will foster widely distributed collaborative learning within engineering teams as vibe coding tools improve, potentially collapsing traditional role boundaries.

Industry leaders agree that AI coding assistants bring builders closer together, with engineers owning core infrastructure while PMs and designers take on more hands-on prototypical work as tools mature.

A Meta product manager credits AI-powered vibe coding tools with transforming his work since mid-2024, enabling him to act more like a builder and product owner.

Arnovitz describes AI coding as giving him “superpowers,” moving to vibe coding after joining Meta in September and using tools like Cursor and models from major providers to ideate, write and review code, and update documentation.

He began using AI coding tools in mid-2024 to generate code from plain-language prompts, enabling rapid prototyping, testing ideas, and improving documentation.