Revolutionising Ansible: Testing the Limits of OpenAI’s ChatGPT for Smarter Automation

OpenAI ChatGPT

ChatGPT is a revolutionary artificial intelligence in that we can interact in a conversational way. Version 3 was launched in November 2022, and ChatGPT in March 2023. Its closest competitor is Google Bard.

See also: Bard-ing with Ansible: Streamlining Testing for Google's AI Writing Tool

Links

• Introducing ChatGPT • GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses

Four Challenges

I decided to judge the quality of the AI using the following four challenges: How to Pass Variables to Ansible Playbook in the command line? - Ansible extra variables Configure a Windows Host for Ansible - Ansible winrm Using Date, Time, and Timestamp in Ansible Playbook - Ansible Tip and Tricks Change user password - Ansible module user

See also: Ansible ansible.builtin vs ansible.legacy: Collection Namespaces Explained

Challenge 1

• How to Pass Variables to Ansible Playbook in the command line? - Ansible extra variables

Score: Good

Challenge 2

• Configure a Windows Host for Ansible - Ansible winrm

Score: Bad

See also: Ansible Multi-Line Strings: Literal (|) & Folded (>) Block Scalars Guide

Challenge 3

• Using Date, Time, and Timestamp in Ansible Playbook - Ansible Tip and Tricks

Score: Good

Challenge 4

• Change user password - Ansible module user

Score: Good

Conclusion

ChatGPT is a great companion that speeds up prototyping and boosts our productivity. I'm impressed by the quality of the result. They appear good-looking and well-organized. However, at the moment, the outcome requires some manual rework before being able actually to use in our laboratory. The challenges saw that it is excellent for minor problems but lacks for those requiring more context interpretation.

Related Articles

• the Ansible Windows reference