Compare international cases before proposing local actions
Ground a local challenge first, then use AI to compare exact case sources before people design and test locally appropriate actions.
Why
Why it earns its place
Outside cases can widen the reference set, but only a traceable comparison shows whether mechanisms and conditions transfer to the local challenge.
How
Run it
Ground the local problem
Humans document local observations, affected people, existing responses, constraints, and unknowns before searching elsewhere. Keep personal and sensitive location data out.
Build an exact case pack
People find three to five international cases and save each URL, date, exact excerpt, context, intervention, and reported outcome under a stable ID.
Compare without invention
Ask AI to compare only the supplied case pack. Require enabling conditions, evidence limits, transfer risks, contradictions, and unanswered local questions.
Propose and test locally
Humans turn mechanisms—not copied solutions—into local actions, mark each inspiration, consult affected people, and test the riskiest assumption first.
Ready to paste
Prompt
Compare a fixed international case-study pack against one local challenge. Do not search for or invent additional cases.
Local challenge: [insert]
Local evidence and constraints: [L1, L2, L3]
Affected people: [generic, non-identifying description]
Case pack: [C1: URL, date, exact excerpts; C2: ...]
Return a table: case ID | problem context | intervention mechanism | enabling conditions | reported outcome and who reported it | evidence limitation | similarity to local evidence IDs | key difference | transfer risk | local question to investigate.
Use only supplied text. Distinguish measured outcomes from author claims and proposed actions. If a field is absent, write “not reported.” Do not infer causation, fabricate success, or recommend copying a case. End with three mechanism patterns humans could examine and the evidence each would still need locally.- Which reported outcome is least verifiable from the supplied excerpt?
- Which local constraint makes an otherwise similar case unsafe to transfer?
The AIdea layer
The designer’s call
- Human core
- Local people define the problem, judge transferability, propose actions, and test them; external cases cannot speak for local needs.
- The trade
- A wider reference set ↔ seductive copy-and-paste solutions. Compare mechanisms and enabling conditions before adapting anything.
- Skip AI when
- Skip AI when exact case sources cannot be supplied or when local evidence and affected-person involvement are missing.
Check before use
Watch out
Remove names and sensitive community data. AI may fabricate case details, outcomes, or transferability; require exact URLs and excerpts, mark “not reported,” and verify every comparison before proposing a local action.
Source of truth
Evidence
A Moldova.org advertorial names one course team combining design thinking, AI, and international cases for polluted-lake actions; it provides no workflow, age-specific result, or outcome measure.
Advertorial · Orange Digital Center: în 2 luni, absolvenții cursului Design Thinking au dezvoltat 8 proiecte ce soluționează problemele locale și globale prin inovație