Comment by wiseowise
15 hours ago
> AI changed all of that. My low-effort issues were becoming low-effort pull requests, with AI doing both sides of the work. My poor Claude had produced a nonsense issue causing the contributor's poor Claude to produce a nonsense solution. The thing is, my shitty AI issue was providing value.
Seems like shitty AI issue did more harm than good?
As I understood it, the "issues" were more like todo list items of "look into whether this is an actual problem" than "this should be fixed".
Yea I think the author is wrong as well. I have a similar skill but the key difference is the instructions are just to fix typos. Why would the author not just use Claude as the plumbing and retain his old nonsense issues is beyond me.
This is the bit that struck me as odd. The author is creating issue slop but blames the contributor for treating it as genuine. The author wants to continue creating slop issues and decides that blocking all external contributions is the solution, rather than spending less time creating slop.
Their slop issues do not actually have value because the fixes based on the slop are equal in their sloppiness.
Author could instead create these slop issues in a place where external contributors can't see them instead of shitting on the contributors for not reading their mind.
Really bizarre lack of self awareness. How do the internal contributors deal with the slop? I wonder what they say about this person in private.
Yeah, I'm baffled that "CEO creates low-effort bug report" -> "open source contributor ignores the low quality of that report and nonetheless fixes the issue in his company's product" is what he apparently considered a healthy open source workflow prior.
That's not his previous workflow. The previous workflow was:
"CEO creates low-effort bug report" -> "CEO uses the low-effort bug report as a starting point to further refine the report and eventually fix the issue in his company's product"
The author's fixes based on the slop are good, because he knows the issue is slop and therefore can improve and fix the sloppiness.
Ignoring AI for a moment: I don't expect anyone to be able to write a design-doc from my own random notes about a problem. They are semi-formed, disconnected ideas that need a lot of refinement. I know that and I have plans around them and know much more context, but if some random person were to take them the outcome would be very bad, or at least require a lot more effort.
A random person has very little chance of being successful with that.
This issue is very similar, only with some AI tools intermediating the notes.