AI-ready questions.
213 ai-readyquestions from the bank — open to read. Pick one and practice it out loud; a coach note comes back in seconds.
Learn the ideas first
All 213 questions
Walk me through the last week of your work. Where did you actually reach for an AI tool, and where did you choose not to?Tell me about a time you caught an AI tool getting something wrong. How did you notice?What's a workflow you've redesigned around AI in the last six months, and what did the old version look like?When you share AI-assisted work with a teammate or manager, what do you make sure they know?What's a task you used to use AI for that you've since pulled back from doing with it?Which AI tools have actually stuck in your workflow, and which ones did you try and abandon?How do you set yourself up to catch an AI mistake before it goes downstream to someone else?How are you keeping up with the pace of AI tools in your field without drowning in it?When is the human still doing the work, even if the AI is producing the output?Have you had to explain an AI-assisted output to a skeptical colleague? How did you handle it?Pick one task you do regularly. Walk me through how AI fits into that task today, end to end.When you've handed something to an AI tool, what's the first thing you look at when it gives the answer back?Describe a time an AI tool was confidently wrong and you had to clean up after it.What's a part of your job that you'd never want AI involved in, and why not?How have your prompts changed over the last year? What were you doing wrong at the start?Have you built or modified a tool, prompt template, or workflow others on your team use? Walk me through it.What signals tell you a task is a good fit for AI, before you've started it?How do you talk about AI use with people on your team who are skeptical or resistant?Tell me about a time using AI made you slower instead of faster. What happened?What's your default tool for writing right now, and what made it the default?How do you decide when to invest in a more elaborate AI workflow versus just doing the thing yourself?Tell me about a piece of AI advice or pattern you copied that didn't work for you, and what you did instead.What does your AI usage look like for the messy, half-formed parts of your work — not the polished outputs?When you share AI-assisted output externally — to a client, customer, or auditor — what changes about how you check it?What's a strong opinion you have about how people on your team should and shouldn't use AI?How would you triage a situation where someone on your team shipped AI-generated content with a factual error?Walk me through how you'd onboard a brand-new teammate to the AI tools your team relies on.If a teammate said 'AI just doesn't work for what I do', what would you actually want to know before agreeing or disagreeing?What's an AI capability that didn't exist a year ago that you've integrated into how you work?How do you handle the question 'did AI write this?' when you've used AI as part of the work but the answer is more nuanced than yes or no?What does a great AI-assisted day look like for you? What does a frustrating one look like?What's a specific case where you'd want a human in the loop even if the AI's accuracy is already higher than the human's?How do you decide whether to fix an AI's output, redo it, or throw it away and start from scratch?What's a non-obvious skill you've found yourself developing because of how often you work with AI?How do you tell the difference between an AI tool that's saving you time and one that's just making your work feel different?When a new AI tool launches in your space, what's your usual evaluation routine?If your manager asked you to teach a 30-minute session on how you use AI today, what would you cover and what would you leave out?What's a use of AI you've seen elsewhere that you think is a bad idea, and what would you do instead?Describe a situation where you had to decide whether to build a custom solution or use an AI tool for a recurring problem. What factors drove your decision?Tell me about a time when using an AI tool actually slowed you down or added friction to your process. What did you learn?How do you decide what level of detail to include in documentation when AI helped you produce the work?Walk me through how you would onboard a new team member on the AI tools your team relies on. What's essential versus optional?Describe a time when you had to validate AI-generated code or analysis before putting it into production. What was your process?Tell me about a time when a client or stakeholder asked if you used AI on their project. How did you respond?What's one skill you've deliberately invested in learning more deeply because AI now handles the basics?Describe a situation where you chose to do something manually even though an AI tool could have done it faster. Why?Tell me about a time when you had to troubleshoot why an AI tool was giving you inconsistent or unexpected results.How do you balance experimenting with new AI tools against the risk of disrupting a workflow that already works?Walk me through a time when you customized or fine-tuned an AI tool's prompts or settings to get better results for your specific use case.Tell me about a time when you had to explain to a non-technical stakeholder how AI was being used in a project and what its limitations were.Describe your process for deciding which parts of a complex task to delegate to AI versus handle yourself.Tell me about a time when you identified a gap in your AI literacy and took action to close it. What prompted that realization?How do you handle situations where team members have very different comfort levels or philosophies about using AI tools?Describe a time when you caught yourself over-relying on an AI tool. What made you realize it, and what did you change?What's your approach to version control or tracking when AI is iterating on your work? How do you stay oriented?Tell me about a project where the use of AI tools raised ethical or compliance questions. How did you navigate that?How do you test whether an AI-generated output is actually meeting the standard you need before you use it?Walk me through how you stay current on which AI capabilities are hype versus genuinely useful for your role.Describe a time when you had to roll back or undo work that was done with AI assistance because it didn't meet requirements.Tell me about a conversation where you had to set expectations with your manager about what AI could and couldn't do for your team.What's a type of work where you've seen colleagues successfully use AI, but you've chosen a different approach? Why?How do you decide when an AI tool's output is 'good enough' versus when it needs more human refinement?Describe how you've integrated AI tools into your daily routine without letting them dictate your priorities or attention.Tell me about the most autonomous task you've handed to an AI agent — something that ran several steps on its own without you in between. How did you decide it was safe to let it go?When you let an AI agent run a multi-step task, where do you put your checkpoints — and why there and not elsewhere?Describe a time an AI agent went off the rails partway through a task. How far did it get before you caught it, and what did the cleanup look like?What kinds of tasks do you now hand to an agent end to end that, a year ago, you'd have done one step at a time yourself?How do you decide between writing one careful prompt and setting up an agent that can use tools and iterate on its own?When an agent can act on real systems — send messages, edit files, run code, move money — how do you scope what it's actually allowed to touch?What's your routine for reviewing the trace of what an agent actually did, versus just looking at whether the final output is right?Tell me about a time you connected an AI tool to your own data or systems — a custom project, a knowledge base, an integration, an MCP server. What did you have to get right?How do you keep an agent on track when the task is ambiguous and it keeps making reasonable-but-wrong assumptions?What's the failure mode you worry about most when an agent can take actions on its own, and how do you actually guard against it?How do you decide how much time, money, or tokens to let an agent burn attempting something before you step in and do it yourself?Walk me through how you'd set up an agent to handle a recurring task for your team. What guardrails would you build in from day one?When have you decided an agent was the wrong tool for something, even though it technically could have done it?How has working with agents changed what 'reviewing the work' means for you day to day?What do you do differently when you can't see an agent's intermediate steps versus when the whole trace is visible to you?Tell me about a time you had to figure out why an agent took the particular path it did. How did you trace it back?How do you handle an agent or tool that succeeds most of the time but fails unpredictably on a small slice of cases?What's the most complex AI workflow you've built or relied on, and what actually held it together?How do you build enough trust in an agent's output to eventually let it run unsupervised on something that matters?When an agent pauses to ask you for confirmation mid-task, how do you decide whether to approve, redirect, or take over entirely?How has the way you break down a task changed now that an agent can carry several steps on its own?If you were handing an important agentic workflow to a teammate to own, what would you insist they understand before they trust it?How do you avoid the trap of building an elaborate agent for something a five-minute manual task would have handled fine?Describe how you keep yourself in the loop on an agent's progress without babysitting every step.What's a task you tried to automate with an agent that just didn't stick? Why not?As agents take on more, how do you keep accountability clear — who owns the outcome when an agent did most of the work?Where do you draw the line on what company or client information you'll put into an AI tool, and how did you arrive at that line?Tell me about a time you wanted to use AI for something but the sensitivity of the data made you stop. What did you do instead?How do you handle confidential, regulated, or customer data when an AI tool would genuinely help but the rules are strict?What's your actual understanding of where your prompts and data go when you use the AI tools in your workflow?Has your organization's AI policy ever clashed with a tool you wanted to use? How did you handle it?What would you do if you found a teammate routinely pasting sensitive data into a consumer AI tool?How do you think about the risk that an AI tool leaks, retains, or trains on what you put into it?What is 'shadow AI' on a team, and how would you handle it if you discovered it was widespread?How do you balance the productivity AI tools give you against the data-governance rules you're required to follow?If untrusted text — a customer email, a web page, a document — feeds into an AI step in your workflow, how do you guard against it manipulating the output?If you were asked to write your team's AI usage policy, what are the first three rules you'd put in, and what would you deliberately leave out?How do you keep a team compliant with AI rules without killing the experimentation that makes the tools worth having?What's your stance on AI output and intellectual property — who owns it, and what can go wrong?In your field, what's a use of AI that's tempting and effective but you'd consider it a compliance or liability risk to actually do?How do you check whether a given AI tool is approved and safe to use before you start relying on it?Tell me about a time you had to say no to an AI use that others were excited about, on risk or governance grounds.How do you handle the audit trail when AI is part of producing a decision or a record that might be reviewed later?How would you respond if leadership wanted to deploy AI in a decision that affects people's outcomes — credit, hiring, care, benefits — faster than you thought was safe?What's a piece of AI hygiene — a habit or a check — that you wish more people on your team practiced?How do you keep AI from quietly introducing someone else's copyrighted or licensed material into your work?When client confidentiality is on the line, how do you decide whether an AI tool can touch the engagement at all?How do you think about bias or fairness in the AI tools you use to make or support decisions?What's your read on the gap between what's technically allowed with AI in your org and what's actually wise? Where do they diverge?What norms does your team have about disclosing AI use, and do you actually think they're the right ones?How should AI-assisted work be credited or attributed when a team ships it together?Have you been in a review where AI-generated work clearly changed the dynamic — what to praise, what to nitpick? What happened?How do you give feedback on work a colleague obviously generated with AI but didn't bother to refine?How do you handle it when someone presents AI output as their own careful, original work?What's the right etiquette for using AI live in a meeting — drafting, summarizing, or fact-checking in real time?How do you mentor or develop junior people when AI can already do much of what they used to learn by doing?What worries you, if anything, about early-career people leaning on AI before they've built the underlying skill?How do you keep a shared standard of quality when everyone on the team is using different AI tools in different ways?How do you handle the productivity-comparison problem — when AI makes it look like one person is shipping far more than another?When you hand off AI-assisted work to someone downstream, what do you make sure travels with it?How do you build a team culture where people are honest about AI mistakes instead of hiding them?How has pair-working or collaborating changed for you when one or both people have an AI assistant in the loop?If two teammates disagree on how much to trust an AI tool for shared work, how do you get them to a workable norm?When you review something a teammate made with AI, what do you spend your attention on now versus two years ago?How do you keep AI from flattening everyone's work into the same voice or the same default approach?What's a healthy team habit around AI you've either started or wish you could start?How do you make sure AI tools raise the floor for the whole team rather than just widening the gap between power users and everyone else?How do you decide which AI model or tool to reach for now that several exist with genuinely different strengths?When is a slower, more expensive, more capable model worth it, and when is the cheap fast one clearly fine?The 'best' tool for your work keeps changing every few months. How do you make decisions without constantly chasing the newest thing?How do you separate real AI-driven productivity from just feeling busy and modern with AI?Leadership says 'use AI to do more with less.' How do you respond honestly without either over-promising or stonewalling?Where in your field do you think it's genuinely irresponsible to use AI right now, even though it's allowed and would save time?What's a task where AI gives you a fast answer that's almost right, and why is 'almost right' actually dangerous there?How do you avoid letting an AI's confident, fluent answer talk you out of your own correct judgment?When AI makes it cheap to generate ten options, how do you avoid drowning in choices and actually decide?How do you tell whether you genuinely understand AI-assisted work well enough to defend it, versus just trusting it looks right?What's a place where you've deliberately kept a human slower-but-accountable instead of an AI faster-but-opaque, and why?How do you decide whether to trust an AI's answer in a domain where you can't easily check it yourself?When AI handles the easy 80% of a task instantly, how do you make sure the hard 20% still gets your full attention?What's an AI shortcut that's popular in your field that you think is actually a bad trade? What's the hidden cost?How do you decide when a problem deserves a custom AI setup versus an off-the-shelf tool versus no AI at all?How do you keep your own skills sharp on tasks you now mostly delegate to AI, so your judgment doesn't atrophy?What's a decision you made recently where AI gave you input, and you went against it? How did you weigh that?How do you know when you've used AI enough on a task and adding more would just be polishing or procrastinating?As AI capability jumps every few months, how do you avoid both over-trusting it today and under-using it out of old habit?What's a task you assumed AI couldn't help with that turned out to be a great fit — what changed your mind?When an AI tool and a trusted human expert disagree, how do you decide whom to believe?How do you guard against AI making you faster at producing the wrong thing — speed without direction?What does your AI setup actually look like right now — which tools, for which parts of your day?What's something you automated in the last month that you used to do entirely by hand?How do you use AI for the thinking parts of your work — framing, deciding, sense-making — not just the producing parts?Show me your most-used prompt, template, or workflow and explain why it's earned a permanent place.How do you use voice, screenshots, or images with AI tools, if at all? What did that unlock?What's a small AI habit that's quietly made a big difference in your work?How do you keep a library of prompts, contexts, or reusable setups so you're not rebuilding them every time?Walk me through how you give an AI tool the context it needs to be useful on your specific work, not generic.What's a task you do so often that you've turned your AI approach to it into a repeatable routine?How do you decide whether to use a general AI assistant or a specialized tool built for your exact task?What's a way you use AI that a colleague was surprised by — something not in the obvious playbook?How do you handle the context limits of AI tools when your real task is bigger than what fits in one go?How do you keep track of versions and changes when AI is rapidly iterating on a piece of work with you?What's your approach to getting an AI tool to match a specific style, standard, or format you need repeatedly?How do you use AI when you're stuck or facing a blank page, versus when you already know what you want?Tell me about a time you chained two or more AI tools together to get something done. What was the seam between them?What's a task where you've stopped writing prompts from scratch and instead built something reusable? Walk me through it.How do you decide how much to invest in setting up an AI tool well versus just using it out of the box?What's the AI tool or feature you reach for most, and what would actually break in your week if it disappeared tomorrow?How do you fold AI into a workflow that involves other people or systems without creating a mess for them?When AI generates a large chunk of work in one shot — too much to read line by line — how do you decide what to actually verify?How do you review work the AI produced that you didn't write yourself, piece by piece? What's your method?AI lets you produce far more than before. How do you keep the volume from outrunning your ability to stand behind it?How do you avoid rubber-stamping AI output just because it looks polished and confident?What's your method for catching the subtle AI errors — the plausible, well-formatted ones — rather than the obvious howlers?How do you handle citations or sources when an AI gives you facts you can't immediately verify?Have you been burned by an AI fabricating something that sounded completely real? What did it cost, and what changed after?Have you ever built a repeatable way to test whether an AI's output is actually good — an eval, a checklist, a test set? Walk me through it.How do you measure whether an AI tool is genuinely producing good output beyond it just feeling right?When AI output is mostly right but has one quiet error buried in it, how do you make sure that error doesn't slip through?What's your tell that an AI answer is probably wrong, even before you've checked it?How do you keep a record of AI mistakes you've caught so the same class of error doesn't keep getting through?When you catch an AI being wrong, how do you figure out whether it's a one-off or a pattern you need to design around?How would you set up a check so that AI-assisted work on your team gets caught before it reaches a customer, not after?What do you do when you suspect an AI output is wrong but you can't quite prove it yet?How has the way you double-check work changed now that more of your inputs come from AI rather than from you?How do you talk to a customer or client about AI being part of the work or product they're paying for?How do you reassure someone worried that 'the AI did the work I'm paying a human for'?How do you set realistic expectations when an executive is over-excited about what AI can do for your team?How do you push back when leadership wants to deploy AI somewhere you think it's premature or risky?How do you explain to a non-technical stakeholder why an AI tool got something wrong, without either excusing it or overcomplicating it?How would you prove to a skeptical leader that AI is actually making your team more effective — not just busier?When someone asks you to 'just have AI do it,' how do you respond if the task isn't a good fit?How do you communicate the limits and uncertainty of an AI-assisted result to people who'll make decisions on it?How do you handle a stakeholder who's afraid AI is going to take their job, when it comes up in your work?How do you communicate an AI policy or new AI workflow to a team that's split between enthusiasts and skeptics?When you present AI-assisted analysis, how do you make clear which parts are machine-generated and which are your own judgment?How do you respond when a stakeholder dismisses solid AI-assisted work simply because AI was involved?How do you keep stakeholders from assuming AI made something faster or cheaper than it actually did?In a regulated or high-trust setting, how do you disclose AI involvement to the people affected by the output?As AI gets better at the core of your job, where are you staking your value — what do you bring that the tool doesn't?What part of your skill set do you think AI makes more valuable, not less?Honestly, which parts of your job do you think AI could automate soon — and how are you getting ahead of that?What's a skill you're deliberately NOT outsourcing to AI because you want to keep it sharp?How has 'getting better at your job' changed for you in an era where the tools change faster than you can master them?What's the most useful thing you've learned about working with AI in the last few months, and how did you learn it?How do you cut through the noise to tell which new AI capabilities are hype and which actually matter for your work?What's a habit you've built to keep learning AI tools without it eating all your time?How do you help your team level up on AI without turning it into yet another mandate that people resent?What's something you believed about AI a year ago that you've since changed your mind on?If you joined a new team tomorrow, how would you get up to speed on the AI tools and norms they already use?