This blog post discusses the dual nature of large language models (LLMs) as both useful tools and products of intellectual labor theft. The author critiques the common pitfalls of LLM usage in software development, urges developers to engage hands-on with coding, and emphasizes the importance of human accountability over AI-generated code. They advocate for public transparency regarding training data and express doubts about open models competing with large labs. The piece concludes with a call to maintain hands-on coding skills to avoid over-dependence on AI technologies.