The problem with this is that you lose the entire context around the answer you're given. There might be notes about limitations or edge cases in the documentation or on StackOverflow. I already see this with my coworkers who use LLMs, they often have no idea how or why things work and then they're absolutely perplexed when something goes wrong.