Lol, it couldn’t determine the right amount of letters in the word strawberry using its training before. I’m not criticizing the training data. I’m criticizing a tool and its output.

It’s amusing to me that at first it’s “don’t blame the tool when it’s misused” and now it’s “the tool is smarter than any individual dev”. So which is it? Is it impossible to misuse this tool because it’s standing atop the shoulders of giants? Or is it something that has to be used with care and discretion and whose bad outputs can be blamed upon the individual coders who use it poorly?