How and why I attribute LLM-derived code | Dark Hacker News