A Cursor developer says engineers need to set 'clear expectations' as AI lets product managers build prototypes

In an evolving landscape where product managers are leveraging AI-powered coding tools to create prototypes, engineers are tasked with the challenge of making these prototypes production-ready. Eric Zakariasson, an engineer at Cursor with a focus on developer experience and product, emphasized the importance of establishing 'clear expectations' between engineering and product teams during the AI Engineer Europe 2026 conference. Zakariasson highlighted the need for clarity regarding what engineers expect from product organizations, suggesting that fully functional apps may not always be the most efficient outcome. With the advent of AI-assisted coding tools, product managers can now produce interactive prototypes without delving into backend systems. These prototypes serve as a demonstration of potential product behavior, illustrating user interactions like button clicks or form submissions. Despite the capability to create these prototypes, Zakariasson noted that they don't need to be fully operational but should provide enough context for engineers to understand the intended functionality. As the line between product and engineering roles becomes increasingly blurred, product managers are taking on the role of 'AI builders', a term that has gained traction among some Meta product managers. This shift reflects a broader movement within companies to empower individuals with AI tools, as illustrated by Meta CEO Mark Zuckerberg's remarks during the company's Q4 2025 earnings call, where he discussed the transformative impact of AI tools on productivity. In line with this trend, LinkedIn recently announced its decision to replace its associate product manager program with training that emphasizes coding, design, and comprehensive product development. However, the rise of AI-assisted coding has also led to heightened demands on engineers. Siddhant Khare, a software engineer at ONA, expressed that while AI has enhanced his productivity, it has also altered the nature of his work. He described the shift from traditional engineering roles to a more evaluative position, likening it to being a judge in an endless assembly line of AI-generated code changes.