I read today how some #software engineers, #coders, #developers are being forced to use #AI in their work. They complain it is eroding their skills. That is often the case with almost any tool that is designed for speed rather than quality. The article makes a good case.
My questions are about AI and software defects, specifically:
1. Do AI tools contribute to better quality software requirements in any way?
2. Does using #AI add defects to the code? If yes, more or fewer than a skilled engineer? More or fewer defects than reused code of any type?
3. Is AI being used to select, plan, or execute the suite of testing procedures appropriate for the software (at any level, i.e., component, unit, or system)? I know this can be a challenge, even for skilled test engineers, because there are so many kinds of tests and tools involved.
Defects can be added or found at any stage of software development. No software of more than trivial complexity is defect free, but good software engineering practice helps drive them out. The earlier they are found, the easier and cheaper they are to correct. The cost of fixing defects late in development can be 100x greater.
https://www.404media.co/software-developers-say-ai-is-rotting-their-brains/?ref=daily-stories-newsletter