I believe the
OpenAI Codex section may be outdated, and it might even warrant being a muted section. However, I’d like to use this opportunity to seek support from the community. As a heavy user of Shell Scripts, I feel that this topic is often overlooked. My hope is that by raising awareness, we can convince the team at OpenAI to improve ChatGPT’s capabilities when handling Shell Scripting languages, such as Bash, ZSH, PowerShell, and others. This way, we can reduce the AI’s hallucinations and increase our trust in its suggestions. Currently, I find myself needing to double-check its accuracy all the time.
The crux of my argument is that regardless of the programming language being used, nearly every programmer has to handle Shell Scripting at some point. This means that Shell Scripting has a broader user base than many other languages, even though it’s “just” a scripting language. This is exactly why AI should excel at it, allowing us to delegate our scripting tasks more efficiently and confidently.