As someone who does have to (sometimes) write the automated tests for my code, here's the tricky part that actually applies to pretty much every use of GenAI:
If you have the knowledge to write the tests yourself, you will be safer to do it.
Because if you don't, you don't know how good they are.
If you have the knowledge to write the tests yourself, you will be safer to do it.
Because if you don't, you don't know how good they are.
Comments
If the check itself is written poorly, it will be useless or even wrong.
And as long as this holds true, the tests will pass.
And if it won't... what can you do about it?
Worse yet, will you even *know* that you need to do something? You didn't write the tests, after all.
And you can't just fire the GenAI or train it like you can with a human.