Thanks Bob, for the rundown. I've tried several of these tools myself and there is clearly a lot of overlap between them. I find myself switching back and forth when the AI gets stuck, which turns out to be useful, but feels a little like throwing shit on different wall to see what sticks.
I have also tried asking one AI to evaluate the work of another, which has yielded some interesting results, although I am still a little skeptical of the output since whenever I propose a refactoring I get a reply starting with "that's an excellent idea", which, statistically speaking, obviously cannot be the case... Some of my ideas must not be great and I would appreciate some constructive critical feedback. Perhaps I need to work on my prompts to facilitate this?
Thanks Bob, for the rundown. I've tried several of these tools myself and there is clearly a lot of overlap between them. I find myself switching back and forth when the AI gets stuck, which turns out to be useful, but feels a little like throwing shit on different wall to see what sticks.
I have also tried asking one AI to evaluate the work of another, which has yielded some interesting results, although I am still a little skeptical of the output since whenever I propose a refactoring I get a reply starting with "that's an excellent idea", which, statistically speaking, obviously cannot be the case... Some of my ideas must not be great and I would appreciate some constructive critical feedback. Perhaps I need to work on my prompts to facilitate this?
Yes! Prompting is an important tool in the process. I'll share my generic instructions prompt which has worked well for me.