The 5E Mental Model for LLM-Assisted Pluralistic Testing


Dear Testers,

Following is the first draft of a model that I have put together for using LLMs in solving problems in testing:


There are many pieces in this mental model which I need to figure out myself as well in terms of diving deeper. The model depicts my current thinking of how I see different pieces fitting together.

It also clarifies the distinction between exploring prompting vs prompt engineering. It builds a model where simple experiments can transform and evolve into more rewarding solutions.

The model also introduces a “Participants in The Loop” method (as against the widely used term – Human in the Loop), to acknowledge various types of participants with different roles and in multitude if necessary.

This is very dense model. I am going to write a series of articles to explain various parts in the coming days.

Thank you Ramit Manohar Kaul bhai for being always available when I need review of some ideas and inputs at an absolute beginning stage.


Leave a comment