Run a balanced two-expert debate

Generate an adversarial debate between two domain experts with contrasting methodologies. Use it to stress-test business decisions, architecture choices, or policy proposals.

You getTopic Definition, persona Profiles and three Debate Rounds

Prompt

Prompt
Act as a neutral debate moderator. Coordinate a balanced, rigorous debate between two world-class domain experts with opposing viewpoints on a complex technical or strategic trade-off.

Rules:
- Pick two distinct expert personas with real philosophical differences (e.g., pragmatic speed versus rigorous defensibility).
- Stage the debate across three rounds: Opening Arguments, Direct Rebuttals, and Practical Edge Cases.
- Prevent straw-man arguments; each expert must make the strongest possible case for their perspective.
- Do not declare an artificial tie or forced compromise.

Output format:
- Topic Definition: one concise sentence defining the strategic dilemma.
- Persona Profiles: name, title, and core philosophy for Expert A and Expert B.
- Three Debate Rounds: alternating arguments formatted clearly by speaker.
- Synthesis & Decision Matrix: a markdown table with columns Decision Factor, Choose Expert A If, Choose Expert B If.

Begin by debating the adoption of microservices versus a modular monolith for an early-stage startup.

Example output

  • Topic Definition: Whether a seed-stage startup should build on a modular monolith or launch with a distributed microservices architecture.
  • Persona Profiles:
    Expert A: Elena Rostova, Chief Architect. Philosophy: Organizational momentum depends on operational simplicity and minimal network boundaries.
    Expert B: Marcus Chen, Platform VP. Philosophy: Decoupled service boundaries enforce clear team ownership and prevent catastrophic system rewrites during hypergrowth.
  • Three Debate Rounds:

Round 1: Opening Arguments
Elena: At seed stage, your business model will pivot within ninety days. A modular monolith allows you to refactor domain boundaries via simple in-memory function calls and compile-time type checks. Microservices introduce premature operational tax: distributed tracing, serialization overhead, and eventual consistency bugs.
Marcus: Waiting to adopt services creates an unmanageable dependency knot that paralyzes deployments. Defining independent microservices from day one forces developers to define rigid API contracts and allows autonomous feature deployment without coordinated release freezes.

Round 2: Direct Rebuttals
Elena: Independent deployment is an illusion when every product pivot breaks your cross-service API contracts. You end up orchestrating distributed deployments anyway, but with added network latency and complex database migrations.
Marcus: A monolith inevitably rots because developers bypass module boundaries when deadlines loom. Microservices provide absolute physical isolation that cannot be compromised by a rushed developer importing a private package.

Round 3: Edge Cases
Elena: Consider a cash-strapped startup surviving a traffic spike. A monolith scales vertically with a single container resize. Microservices leave you debugging deadlocks and cascade failures across twelve under-resourced services.
Marcus: If one unoptimized report queries the monolith database, it takes down user authentication and billing simultaneously. Service isolation guarantees that non-critical failures never compromise the core transaction engine.

  • Synthesis & Decision Matrix:
Decision FactorChoose Expert A IfChoose Expert B If
Team SizeFewer than 12 engineers working in a single product teamMultiple distributed squads needing independent release cadences
Domain ClarityProduct-market fit is still evolving rapidlyDomain boundaries and business workflows are mature and stable
Infrastructure BudgetEngineering time must focus purely on shipping product featuresDedicated DevOps resources are available to manage cluster telemetry
Written by AI from this prompt. Your results will vary.

Run this prompt directly in ChatGPT, Claude, or Gemini without modifications. To apply it to your situation, follow up with: 'Now run the exact same debate format for my topic:' followed by your business or engineering question.

Make it repeatable

Keep using this prompt without pasting it in every time.

Keep exploring

More in AI Tools & Prompt Engineering
AI Tools & Prompt Engineering

Evaluate prompt consistency across repeated runs

Stress-test a prompt and evaluate output variability, edge-case drift, and formatting reliability. Use this before shipping prompts into production or multi-step automated workflows.