Not debating this but an interesting experiment that I ran which worked pretty well was asking ChatGPT to segment off different personalities that would critiqu

Not debating this but an interesting experiment that I ran which worked pretty well was asking ChatGPT to segment off different personalities that would critique the code from different perspectives. Sample Prompt: Pretend you have 2 additional personalities, one is a security engineering expert and computer hacker, the other is a lead software engineer who cares about software maintainability. Call them Bob and Evan. Have them analyze the code from their perspectives and show me their thoughts on the code you've written. That usually generates fairly good critiques of code snippets, examples, etc.