Is Cursor Really That Great? A Veteran Programmer's Honest Review
The author provides an in-depth comparison of the popular code completion tool Cursor with other options. They find that Cursor's core technology is not fundamentally different from Copilot, both relying on Claude or GPT models. Cursor excels in actively searching and referencing other files within a project, but it can sometimes be overly 'smart,' even creating new files without permission. The author prefers the o1 model for its more precise and reliable debugging capabilities. The article concludes that those excessively praising Cursor might lack programming experience, confusing the power of AI with the tool itself. The author stresses that choosing an editor should be based on personal preference rather than blind following of trends.
Read more