Ethics are important. It’s OK to be mad at Apple for using child labor, even if it gets them better financial results. Similarly, it’s OK to reject LLMs for their disproportionate energy use compared to more trustworthy alternatives.
For one example like the one you gave where you seem satisfied about the output, there’s another where the code it spits out is unusable (cf gamedev), and you can refine prompts as much you’d like, but it’ll make you wish you started looking for a solution the old-fashioned way to save some time.
Additionally, do you remember the study that showed that AI-generated code is 40% (not sure about the figure) likely to be buggy? Because I do.
At the end of the day, there’s no need to rush to reach a broken solution. You’re better off not cheating and doing things you can actually understand, audit and maintain.
Ethics are important. It’s OK to be mad at Apple for using child labor, even if it gets them better financial results. Similarly, it’s OK to reject LLMs for their disproportionate energy use compared to more trustworthy alternatives.
For one example like the one you gave where you seem satisfied about the output, there’s another where the code it spits out is unusable (cf gamedev), and you can refine prompts as much you’d like, but it’ll make you wish you started looking for a solution the old-fashioned way to save some time.
Additionally, do you remember the study that showed that AI-generated code is 40% (not sure about the figure) likely to be buggy? Because I do.
At the end of the day, there’s no need to rush to reach a broken solution. You’re better off not cheating and doing things you can actually understand, audit and maintain.