I asked Claude, ChatGPT, and Gemini to fix the same bug, and only one understood it
…Gemini 3.1 got 0 out of 3 Gemini was the most technically verbose out of the three, and also the most off-target. Instead of diagnosing the planted bugs, it flagged…
