header-langage
简体中文
繁體中文
English
Tiếng Việt
한국어
日本語
ภาษาไทย
Türkçe
Scan to Download the APP
Partnered With:

「Clean Code」 and Line-by-Line Review Are Dead, Ramp CEO Declares AI End Traditional Software Development

According to Perceive Beating monitoring, Ramp AI lead Rahul Gupta pointed out that the next generation of large models is transforming programming into a "natural language to code interpreter," fundamentally disrupting traditional software development logic. Gupta warned that with the upcoming models like Anthropic Fable crossing the accuracy of code issues and output complexity, rigid adherence to "clean code" and manual line-by-line review in low-risk areas is no longer necessary. Line-by-line reviews should only be reserved for high-risk core modules involving security and finance, shifting the focus of software development towards systematic empirical validation and embracing technical debt proactively.

In the new development paradigm, AI significantly reduces refactoring costs, and reasonable abstraction and perfect architecture are becoming less critical. Development teams can accept a 50% increase in redundant code for a 5% performance improvement, leaving the accumulated technical debt to be maintained directly by more intelligent models in the future. Gupta noted that the bottleneck of software delivery has now shifted entirely to the "review and merge" phase. For low-risk modules, they should be treated as a "black box" just like neural networks, isolated in a sandbox with all outbound access to databases and networks completely cut off. Thousands of real-world input empirical validations should be performed in the CI process to replace inefficient manual line-by-line logic verification.

Boris Cherny, the father of Claude Code, strongly concurred with Gupta's disruptive remarks, adding that since models can now correctly generate most task code, human core work is transitioning to designing a secure defense network for the closed-loop operation of large models and validators, continuously eliminating various system bottlenecks in the engineering process.

举报 Correction/Report
Correction/Report
Submit
Add Library
Visible to myself only
Public
Save
Choose Library
Add Library
Cancel
Finish