Interview bot beta

Ask better questions

Use this AI Interview Lab to practice refining broad inquiries into specific, accountable reporting questions. Step into the story, question your sources and hone your reporting style.

→ Download the .md file.
→ Drop it into your favorite LLM.
→ Prompt, "I want to play this game."
Save interviewbot.md
Bowling header image

How this interview lab works

You’ll jump into a bowling story with three built-in interviews: two competitors and one outside voice. Everyone gets the same setup, so it’s your reporting instincts that make the difference.

How many questions you can ask

  • You start with up to three questions per source.
  • You can earn one bonus question if your first three are strong and none are Tier 1.
  • The maximum is four questions for one source.
  • The interview stops when you hit the question cap.

Professional reporting standards

  • Do not invent facts when you challenge a source.
  • Clearly separate confirmed facts from rumor-level information.
  • Do not use coercive tactics or pressure for private health details.
  • Represent earlier answers fairly in follow-up questions.

How question tiers work

  • Tier 1: Broad or generic prompts with little reporting value.
  • Tier 2-3: Clear topic focus but limited evidence or accountability.
  • Tier 4-5: Specific fact-based questions that test decisions and impact.
  • Your goal is to consistently ask Tier 4 and Tier 5 questions.

Question quality guide

Use these patterns to see the difference between weak, decent and strong reporting questions.

School budget hearing

Tier 1 question

"Any comment?"

Tier 3 question

"Are you worried about the budget cuts?"

Tier 5 question

"You voted to cut after-school transit while keeping athletic travel unchanged. What criteria produced that tradeoff and what impact do you expect on students who rely on late buses?"

Why stronger: It names a decision, asks for criteria and asks for consequences.

Startup product launch

Tier 1 question

"Was the launch bad?"

Tier 3 question

"How did launch day go?"

Tier 5 question

"You promised a same-day support response but users reported 18-hour waits in the first week. What broke in your support model and what changes are in place before the next release?"

Why stronger: It anchors the question to a measurable claim and demands accountability plus a forward plan.

City transit delay report

Tier 1 question

"Why is transit always a mess?"

Tier 3 question

"What caused the delays today?"

Tier 5 question

"Riders were told there was a signal issue but internal alerts show two staffing gaps and one equipment fault. Which cause created the longest disruption and what can riders expect tomorrow morning?"

Why stronger: It tests an official explanation against specifics and asks for practical next-day impact.

Use this structure in your own reporting: specific evidence, clear accountability and concrete impact.