"Coding is solved" misses the point

The article challenges the recurring claim that 'coding is solved' by AI. It acknowledges that AI can generate code for well-specified problems, but argues that real software engineering involves much more than writing code. It highlights the importance of organizational context—such as existing architecture, security requirements, business constraints, and team ownership—which AI cannot easily grasp. The article suggests that as implementation becomes cheaper, the bottleneck shifts to higher layers like product design and solution design. It concludes that AI makes implementation easier but doesn't solve the broader discipline of software engineering, which requires deep understanding of the organization and its goals.