Do Users Write More Insecure Code with AI Assistants?https://arxiv.org/pdf/2211.03622.pdf?trk=public_post_comment-text