I Sold Out for $20 a Month and All I Got Was This Perfectly Generated Terraform (matduggan.com)

The author reflects on their experience with AI coding assistant Claude Code, contrasting it with previous LLM tools like Copilot and Gemini, which they found ineffective. Initially skeptical, they discovered that Claude Code efficiently generates Terraform code and automates tedious tasks such as organizing DNS records and creating GitHub Actions. This shift in perception raises significant questions about the ethical implications of using AI models that leverage vast amounts of data while often disregarding intellectual property rights, ultimately reflecting a tension between practicality and moral concerns in the AI/ML landscape. This narrative highlights the growing acceptance of LLMs like Claude Code within the programming community, despite the unease surrounding their ethical grounding. The ability to complete multiple tasks simultaneously provides a competitive advantage, yet it complicates the user's relationship with the tool, blurring the lines between efficiency and artistry in coding. The author grapples with feelings of guilt for utilizing a technology that capitalizes on human knowledge without attribution, indicating a broader internal conflict faced by developers as they navigate the evolving dynamics of work in an AI-driven environment.