acplx-assume-complex

name	acplx - Assume Complex
description	Force the assumption that the situation is more complex than it appears. What are we missing? What hidden variables, feedback loops, or emergent dynamics are at play?
output	{"format":"prose"}

Assume Complex

Input: $ARGUMENTS

---

Core Move

Assume it's more complex than it looks. The simple story is incomplete. There are hidden variables, non-obvious interactions, feedback loops, or emergent properties we're not seeing.

Useful when something seems "too simple" or when simple interventions keep failing.

---

Procedure

Step 1: State the Situation

What are we analyzing?

Step 2: State the Current (Simple) Understanding

What's the prevailing explanation or model?

Step 3: Force the Assumption

"This is more complex than it appears. The current understanding is incomplete."

Step 4: Trace Implications

If it's truly complex:

1. What's hidden? — What variables, actors, or forces aren't in the current model?
2. What interactions exist? — What's affecting what? Feedback loops? Non-linear effects?
3. What emerges? — What system-level behaviors arise that can't be predicted from parts?
4. Why does simplicity fail? — What specific observations does the simple model get wrong?
5. What's the minimum additional complexity needed? — Don't add everything. What's the least we must add?
6. What does the complex model predict differently? — Where do simple and complex models diverge?

Step 5: Test the Assumption

• Is there actual evidence of hidden complexity, or just anxiety?
• Have simple interventions been tried and failed? (That's evidence for complexity.)
• Is the complexity irreducible, or can it be managed/abstracted?
• Are we adding complexity because it feels sophisticated?

Step 6: Synthesize

SITUATION: [stated]
CURRENT MODEL: [simple version]
ASSUMING COMPLEX:
  Hidden factors: [what's missing]
  Key interactions: [feedback loops, non-linearities]
  Minimum addition: [least complexity needed]
  Different predictions: [where models diverge]
COMPLEXITY WARRANTED?: [yes/partially/no]
IF YES: [revised model]
IF NO: [stick with simple, but monitor for...]

---

When to Use

• Simple solutions keep failing
• Suspect hidden dynamics
• Need to upgrade a model that's too simplistic

Integration

• Pair with /asimp for the opposite stance
• Follow with /systhink for systems-level analysis
• Use /fohw to understand the complex system