Skip to main content

​
What You’ll Learn

  • How Debate mode works
  • What the debate structure looks like
  • When it’s the right mode
  • How to frame questions for the best debates

​
How It Works

  1. You pose a question, statement, or decision
  2. AIs are assigned different positions (for/against, or multiple viewpoints)
  3. Each AI argues their assigned position with evidence and logic
  4. AIs respond to each other’s arguments (rebuttals)
  5. You see the full debate and decide for yourself
The key: AIs argue positions they’re assigned, not necessarily what they’d “naturally” recommend. This ensures you hear the strongest case for each side.

​
What You See

A structured debate with:
  • Position statements - Each AI’s initial argument
  • Evidence - Data and reasoning supporting each side
  • Rebuttals - AIs responding to each other’s points
  • Key tensions - Where the fundamental disagreements lie
  • Common ground - What both sides agree on (often the most useful insight)

​
When to Use Debate Mode

ScenarioWhy Debate
”Should we do X?” decisionsSee both sides fully argued
Controversial topicsGet balanced perspectives
Before making big commitmentsStress-test your leaning
When you suspect confirmation biasForce yourself to hear the other side
Strategy choices with trade-offsUnderstand what you’re giving up

​
Example

Your message: 
Should we raise our Series A now or wait 6 more months
to improve our metrics?
Debate (excerpt):
FOR: Raise Now (Argued by Grok, GPT-5.2)
  • Current market conditions favor AI companies - window may not last
  • 78% of startups that delay fundraising report worse terms 6 months later
  • Your current metrics ($45K MRR, 15% MoM growth) already meet typical Series A benchmarks
  • Runway anxiety affects team performance and hiring ability
  • Rebuttal to “wait”: The metrics improvement you’re targeting (2x MRR) isn’t guaranteed and may take longer than 6 months
AGAINST: Wait 6 Months (Argued by Claude, Perplexity)
  • Current MRR (45K)putsyouatthelowendofSeriesArange−45K) puts you at the low end of Series A range - 100K+ gets significantly better terms
  • 6 months of growth at 15% MoM = $105K MRR (more than doubles valuation leverage)
  • Fundraising takes 3-4 months - you’d be raising during a stronger position
  • Fewer dilution at higher valuation: potentially 5-8% less founder dilution
  • Rebuttal to “raise now”: Market timing is unpredictable, but metrics-based valuation is more defensible than momentum-based
COMMON GROUND Both sides agree:
  • Current metrics are fundable, just not optimal
  • Market conditions are favorable but uncertain
  • The decision hinges on confidence in sustaining 15% MoM growth
KEY TENSION Risk of waiting (market downturn, growth stall) vs. reward of waiting (better terms, less dilution)

​
How to Frame Questions for Better Debates

​
Good Debate Questions

  • “Should we build or buy our analytics solution?”
  • “Is it better to hire a senior marketer or two junior ones?”
  • “Should we expand to Europe now or focus on US market?”
  • “Remote-first or hybrid: which is better for our 30-person team?”
These work because they have legitimate arguments on both sides.

​
Poor Debate Questions

  • “What’s 2 + 2?” (no debate to be had)
  • “Tell me about marketing” (no clear positions)
  • “How do I fix this bug?” (not a debate topic)
Debate mode shines when there’s genuine tension between options.

​
Tips

  • State your current leaning if you have one: “I’m leaning toward raising now, but change my mind.” This makes the counter-arguments more targeted.
  • Follow up on the strongest counter-argument. The point that surprises you most is probably the one worth exploring.
  • Use Debate before Red Team. Debate is balanced (both sides). Red Team is adversarial (attacking your idea). Start with Debate for perspective, then Red Team to stress-test.
  • Generate a Decision Record after a good debate - it captures the arguments, trade-offs, and your final rationale.
  • Don’t treat it as a vote. 3 AIs arguing “for” doesn’t mean it’s the right choice. Evaluate the quality of arguments, not the count.

​
After the Debate

Once you’ve seen both sides:
  1. Dig deeper on the argument that surprised you: “@Claude, tell me more about the dilution impact you mentioned”
  2. Ask for a recommendation now that both sides are laid out: “Given everything discussed, what would you actually recommend?”
  3. Generate a document - Decision Record captures the full reasoning for future reference

​
Still Need Help?

Reach out to us at [email protected] or use the feedback button in the app.