Fidelity & Additional Scope Control

When trying to replicate work of the perfect human practitioner, I’ve found that due to the variability of what perfection means across practitioners, I needed a way to accommodate the whims of stakeholders - which is what they do. As a result, I’ve begun to add in fine-tuning controls.

These are still under development and testing, but here ya go anyway. Maybe you’ll find some additional things you need to control that I haven’t thought of.

The Prompt

Map the End Users Job-to-be-Done

Below I’m going to itemize the points I use in the prompt and explain them. Then I’ll give you a copy and paste version since I don’t generally write them in separate lines. That will be the last installment of this series.

BEGIN!

Fidelity

An ideal job map will have 10 to 18 steps. If {{fidelity}} = ‘low’ then the number of steps should be closer to 10. If {{fidelity}} = ‘med’ then the number of steps should be closer to 14. If {{fidelity}} = ‘high’ then the number of steps should be closer to 18.

Example:

• Low = 10 - 12 steps
• Med = 13 - 15 steps
• High = 14 - 18 steps

Instructions: I’ve been luck in that I’ve seen a lot of Job Maps that aren’t public. In the array I’ve seen everything from 7 to 30 steps. What I don’t want to do is give AI the discretion to generate the perfect map because it hasn’t had discussions with my stakeholder, and I don’t want to roll the dice all day. So, I’ve constrained it a bit. This is a very low fidelity (no pun intended) way of accomplishing this. I continue to work on this.

MECE

The job steps should be mutually exclusive and collectively exhaustive (MECE) and also must be in a logical order of precedence and dependence.

Think this through step-by-step.

Instruction: I’m not certain that this is necessary (the MECE part), but what the heck? The sequence, however, is an instruction that I feel needs to be here.

Scoping

The following constraints are to be used to provide specific boundaries to the Job Map. Follow them carefully.

• The first step will generally be about {{start point}}. Give the step an appropriate name and format as previously described. Do not begin with anything that would logically precede this.
• The last step will generally be about {{end point}}. Give the step an appropriate name and format as previously described. Do not include any steps that come after this.
• Please follow all instructions carefully.

Instruction: this component allows me to determine where the map begins and where it ends. Theoretically, this overrides portions of the phasing. What it doesn’t do is let me reorder the phasing as we saw in the anatomical structure job. Personally, I see no reason to do ever allow that. I’d rather advise my stakeholder effectively so they understand the research we’re doing.

Variables

These variables appear at the end of the entire prompt - along with the other variables - to provide the control we are seeking. As you can see in the instructions, I use “generally be about {{…” instead of a definitive step. You can enter a definitive step or some type of descriptive phrase. The AI figures it out.

For fidelity, I simply delete two of the values (or all of them).

start point: end point: fidelity: low | med | high