Someone submitted a AI-slop PR to a project I'm working on. So fucking annoying.
Did you not bother checking to see if it created a bunch of slop, or are you incapable of realizing it's slop? Not sure which is worse. And now I have to sit here and code review your slop, which basically is just you pushing your work to me.
Many things, but as an example, catching an error based on the assumed content of the error message ("does the error message contain this word")and re-bubbling the error for no purpose.
"AI" poops out a lot of code that technically "works" but is not scalable and would not pass code review.
Not an error code, the message itself. It was testing independently whether a word in the message was upper case or lower case (Does string contains "Something"? Does string contain "something"). It was just a "guess" on what some random 3rd party library might emit as an error message.
There was also no need to catch and re-throw an error with a different message in this context - just let the error happen, nothing was handling it.
I agree. Most AI problems come from people who don't know how to do their jobs, or use AI. They just assume the AI made it for them. They don't understand the genius of the answering machine that asks if the screen is hooked up, the computer plugged in, and if it is turned on.
Someone submitted a AI-slop PR to a project I'm working on. So fucking annoying.
Did you not bother checking to see if it created a bunch of slop, or are you incapable of realizing it's slop? Not sure which is worse. And now I have to sit here and code review your slop, which basically is just you pushing your work to me.
Many things, but as an example, catching an error based on the assumed content of the error message ("does the error message contain this word")and re-bubbling the error for no purpose.
"AI" poops out a lot of code that technically "works" but is not scalable and would not pass code review.
Not an error code, the message itself. It was testing independently whether a word in the message was upper case or lower case (Does string contains "Something"? Does string contain "something"). It was just a "guess" on what some random 3rd party library might emit as an error message.
There was also no need to catch and re-throw an error with a different message in this context - just let the error happen, nothing was handling it.
I agree. Most AI problems come from people who don't know how to do their jobs, or use AI. They just assume the AI made it for them. They don't understand the genius of the answering machine that asks if the screen is hooked up, the computer plugged in, and if it is turned on.