ARCHIVE

What would make you turn the feature off for 24 hours, no debate?

Do your logs capture user intent, model output, and downstream impact in one chain?

Which edge case appears often enough to be called a normal case now?

Where does your incident response playbook mention model behavior explicitly?

Are you shipping capability, or shipping legibility about capability?

What hallucination is your team quietly tolerating because the UI looks polished?

If a regulator audited your AI feature tomorrow, what evidence would you show first?

Which business decision is currently delegated to a model without explicit ownership?

What percentage of outputs should be refused, and do you treat refusal as success?

Are you training users to trust your system, or to double-check everything manually?

What assumption in your eval set would break first with a new customer segment?

Which human review step is real quality control, and which is theater?

If your retrieval layer disappeared, how much of your AI intelligence remains?

Where are you paying for model tokens to avoid hard product decisions?

Are you measuring latency users feel, or only latency your dashboards can see?

What part of your architecture is optimized for demos instead of Tuesday morning traffic?

If you removed your top three prompts, would your product still work?

Which failure mode would cost your team trust fastest: silent wrong answers or obvious downtime?

Where in your stack are you still confusing model fluency with system reliability?

If your flagship AI workflow had to justify its existence in one KPI, which KPI would survive scrutiny?

What is the uncomfortable truth you already know but have not yet operationalized?

Are you building artifacts that compound, or only performing effort?

What would you stop reading so you can start building?

If your next piece had to be useful to one person, who would it serve?

Which fear is currently editing your output before the first sentence?

What does creative consistency require from your calendar, not your mood?

Are you shipping work that teaches others, or only work that signals taste?

What part of your voice disappears when you imitate professional tone?

Which collaboration would accelerate your growth if you asked directly?

What would happen if you treated drafts as assets instead of evidence of inadequacy?

Are you consuming more creative work than you are producing?

Which idea keeps returning and what does that repetition demand from you?

Are you documenting your process so future you can repeat results?

What would you publish this week if polish were capped at 80%?

Which audience are you trying to impress, and is that audience worth the cost?

What project are you avoiding because beginner-level output feels threatening?

Are you waiting for inspiration to start what a system could finish?

Which constraint makes your creative output better, not smaller?

What part of your work stays unfinished because you are protecting your identity?

Are you creating to discover truth, or creating to manage perception?

What would enough look like in a year you can actually sustain?

Are you building a life system, or improvising the same week repeatedly?

What one change would reduce decision fatigue before noon?

Which promise to yourself has become optional without your consent?

What does a sustainable minimum viable day look like when motivation is zero?

Are you designing your environment for the person you are, or the person you wish you were?

What belief about discipline is actually a disguised perfectionism trap?

Which responsibility should be a checklist instead of a decision?

If your phone disappeared for seven days, what would improve first?

Where are you mistaking intensity for consistency?

What trigger reliably knocks you off track, and what is your preplanned countermeasure?

Are you optimizing for short-term comfort or long-term capacity?

Which boundary, if enforced, would improve both your work and your relationships?

What would your week look like if recovery was treated as production infrastructure?

Which commitment looks admirable publicly but is eroding you privately?

What personal metric would reveal your busy but drifting weeks immediately?

Are your meal choices driven by strategy or by whatever is closest?

If your energy crashes at 3 PM, what system have you refused to redesign?

What part of your schedule is performance, and what part is avoidance?

Which daily habit gives you the highest return and why is it still negotiable?

Are you earning trust session by session, or spending it?

If support had veto power, what would they stop you from launching?

Which product risk are you treating as a future problem even though usage already proves it?

What one constraint would force your roadmap to become honest?

Are you collecting user feedback fast enough to influence this quarter, not next year?

Which integration creates the most support tickets per dollar of revenue?

Are your KPIs aligned with user outcomes or with internal optics?

What does good enough mean numerically for your core user journey?

Where does your UX hide uncertainty instead of communicating it?

Which customer segment gets your best reliability, and which gets your leftovers?

What evidence would convince you to kill a feature you personally love?

Are you shipping options, or shipping clarity?

Which feature would users miss tomorrow, and which would they barely notice?

What would your churned users say you still refuse to hear?

Is your product teaching users better judgment, or dependency on automation?

Which onboarding message sets expectations you cannot meet consistently?

What workflow step do users repeatedly hack around, and what does that reveal?

Are you instrumenting adoption, or just measuring clicks?

Which user pain are you calling behavioral because fixing it is expensive?

If users got one minute with you, what misconception about your product would you correct first?

What decision are you making by default that deserves explicit strategy?

Are you scaling headcount, or scaling decision quality?

Which quick win is quietly creating a long-term maintenance trap?

What standard would immediately improve quality if you enforced it this week?

Where are you overfitting process to one painful incident from the past?

Which tradeoff is obvious technically but politically uncomfortable to say out loud?

What would your team stop doing if it had to defend every recurring ritual?

Are you hiring for credentials, or for evidence of judgment under ambiguity?

Where are your senior people still operating without a repeatable system?

What dependency could block you for weeks if its API changed tomorrow?

Which part of your org chart hides the real decision maker for risk?

Are you rewarding engineers for shipped outcomes or for visible activity?

What signal tells you a project needs a staff engineer, not another sprint?

Which meeting could be replaced by a clearer written technical decision record?

What are you over-centralizing that should be codified as team-level standards?

If budget dropped 40%, what part of your AI stack would you keep?

Which decision are you postponing by saying we need more data?

Are you building faster, or just accumulating unresolved technical debt faster?

What promise does your marketing copy make that your architecture cannot keep?