“ChatGPT answers more than half of software engineering questions incorrectly”
That 40% didn’t spot the incorrect answers is a classic case of automation bias. (Cognitive automation tools help you think less and thinking less makes you less likely to spot errors)