Modeling executive function as a grey-box system: Valid or delusional?

I’ve spent the last year treating executive function attributes (Discipline, Grit, Patience, etc.) as mutable "system variables" instead of fixed traits.

I’m taking a systems-engineering approach to “prescribe” CBT-based psychological interventions when I hit a wall (e.g. procrastination, "redlining"), and also to identify which traits need the most attention to optimize overall “system” (my) output.

*Model Overview*:

I’m using a Grey-Box model, dependent on the following premises:

1. Performance as Function: Performance output is in large part a function of the psychological traits I listed.

2. Mutability: These traits are mutable to the extent that maximizing them results in non-trivially enhanced performance.

3. Control Theory Loop: Defining a self-correcting model for these traits using control theory, benchmarking it vs some ground truth (e.g. features shipped, PnL), & running trial and error on individually-tailored psychological interventions can yield actionable growth insights.

*The Bottlenecks*:

1. It’s vulnerable to heavy GIGO ("Garbage-In-Garbage-Out").

2. Accurate state estimation currently requires high-friction journaling, which is the primary failure point, but not insurmountable w/ the advent of LLM journaling

*Questions for HN*

1. Prior Art: Do any of you take a structured, quantified approach to improving these (or any other) personality traits?

2. The Placebo Check: Please poke holes in the logic. I worry the success I’ve personally found with this approach may just be a placebo.