Uhm no, if an AI writes software that contains mistakes the errors would still be important as now with human-written software. When a programmer makes a mistak

Uhm no, if an AI writes software that contains mistakes the errors would still be important as now with human-written software. When a programmer makes a mistake usually we don't think about intentions but this is because we usually assume the mistake is unintentional. With AI we can ask ourselves if the mistake is intentional or not, but the mistake would still be important.