Ask HN: What’s the Most Challenging or Fascinating AI Project You’ve Developed?

The author prefers technically complex projects, particularly in systems programming, over web applications or simpler tools. They have utilized various tools, including fuse filesystems, inter-language bridges, ORM adapters, framework code, and small platform scripts, often aided by AI for efficiency. A noteworthy aspect of their work is the extensive use of Rust, despite not being proficient in it. They emphasize understanding code blocks and carefully reviewing tests in a familiar language. The main AI tool used is aider-chat. Key tips include adopting Test-Driven Development (TDD), following systematic design and troubleshooting loops, and multitasking with tools like tmux to minimize downtime while waiting for responses from the language model.

Source link