Sorry in advance for the maybe naive-offensive question. I admire you all. Only developers have the knowledge to understand the quality of suggestions or code problems that an ai model finds. I just run a check in a well known AI model (i am not writing to advertise any company) that offers free hundreds of thousands of tokens. In a few minutes x11window.cpp, window.cpp, workspace.cpp, input.cpp and useractions.cpp were checked for free by the most advanced model. I find that first scary and intimidating but also impressive. Nothing grave was found but dozens of suggestion for potential bugs.By any means i am not capable of judging the result. I just think of the hundreds of bug reports that consume valuable time from the developers. Thank you in advance
Please do not report AI bugs.
7 Likes
Let’s not put AI in everything.
2 Likes
AI tools for code are essentially only useful for developers of that code since they can make actually informed decisions on whether the AI output is good or bad.
A more low hanging fruit might be using AI tools to analyze bug reports and find duplicates and/or related reports.
Potentially already during entry of bugs, avoiding duplicates and/or guiding bug reporters.
In any case you’ll always need a human to make the final decision as bugs can sounds similar on a superficial glance but quite different when looked at in detail
4 Likes