Not a research post, just something I've been chewing on after watching this happen a few times now. When a human reviewer leaves a comment saying "this looks like a bug," people push back, ask questions, sometimes just disagree outright. When an AI leaves the exact same comment, phrased almost identically, people tend to just fix it. Same words, different reaction. Took me a while to figure out why that bothered me. It's not that the AI is wrong more often, it's actually pretty accu...