Cursor's new visual editor: Right idea, wrong implementation (blog.nordcraft.com)

Cursor has introduced a new visual editor for its AI coding tool, addressing a significant challenge in front-end development where large language models (LLMs) struggle with ambiguity. While LLMs excel at generating backend code for well-defined tasks, their performance diminishes when specificity is required for user interface design, leading to frustration and multiple iterations for developers. The visual editor allows users to make precise changes to their interfaces, which are then converted into prompts for LLMs, speeding up the coding process and reducing the need for repeated adjustments. However, the implementation raises concerns about its effectiveness. Critics argue that while the aim is to streamline interface design, the requirement for explicit instructions negates the unique advantages of utilizing LLMs, as developers still have to verify the output. Moreover, the processing adds unnecessary costs by consuming tokens for user-generated instructions. In contrast, Nordcraft is poised to offer a more integrated solution by combining LLM capabilities with precision in a proprietary visual editor, aiming to enhance both efficiency and clarity in web development without the pitfalls identified in Cursor's approach. Nordcraft's AI assistant is set for release in early 2026, promising a more seamless coding experience.