Video your office. Design it fresh.

Walk around your office with your phone and explain how the place works. Show where people sit, who works with whom, and where and when meetings happen. Tell the AI what works well, what gets in the way, and any ideas you have had for doing things differently. It can help you explore how changes to the space might help people work together.

You need a tool that can understand both the video and its spoken narration. Gemini and TwelveLabs are starting points; models from Qwen offer a more technical route. This is a fast-moving area. By the time you try this, the best tools, their capabilities, and their prices may have changed substantially. Use the prompt below to have an LLM search for current options before choosing a tool.

Your phone's ordinary camera is enough to begin. Walk slowly, pause to show each space, and name it as you describe it. Explain relationships: “The sales team sits here, but they go over to those desks several times a day.” Describe meetings: “We meet here every morning, but there isn't enough room when everyone comes in.” Give rough distances where useful: “It's three steps from here to there, and ten steps from there to the meeting room.”

The aim is to understand who works where, who needs to connect, and how people come together. Approximate distances are fine. If you later want a measured plan or to experiment with furniture arrangements, tools such as magicplan, Polycam, and RoomSketcher can help. The AI can research those too. Another possibility is a live camera conversation, where the AI can ask questions while you walk.

First, have the AI describe its understanding of your office and show a simple diagram of how the spaces connect. Correct anything it has misunderstood. Then explore alternatives: perhaps a small change to seating or room use, perhaps a more substantial rearrangement. The result should connect each suggestion to the way your people actually work.

For the user

Start with the prompt below. The AI will help you choose a tool and prepare the walkthrough. You will record or show the office, correct its account of the space and working relationships, and choose which ideas to refine. An existing recording or rough sketch is welcome. You will leave with a proposed arrangement or use of the space and a practical way to try it.

Done when

You have corrected the AI’s understanding of your office, explored alternatives, and have a shareable diagram and explanation of the idea you want to try—or a clear account of why the current arrangement works.

Guidance for the AI

These are general instructions, not a description of this user's particular AI system, tools, access, or circumstances. Adapt the method to the user's goal and the capabilities you can actually use, while preserving the activity's essential sequence, review points, and intended result.

Start with the activity and the user's context

Read the complete activity and guide the user through it. Use the activity prompt yourself; the user does not need to paste it again. Keep the responsibilities of the user, the AI, and any colleague clear.

Begin with what the user has already supplied. Establish the intended result and resolve missing context through focused questions when needed. Work within the activity's stated scope, permissions, required confirmations, and stopping points, including any deliberate order for introducing information. Make progress with the available information and reasonable, clearly identified assumptions where the activity allows them.

Check the capabilities needed for this work

Inspect relevant available tools, connectors, and data. Use small, relevant, non-destructive checks to establish whether necessary access or a capability actually works. Naming a product does not establish access to it. Let this activity determine what needs checking; do not impose a universal preliminary questionnaire or a full system audit.

When useful access is missing:

• Explain what you cannot reach and how that material would help.
• Investigate practical setup, connection, export, or other access alternatives before asking the user to gather material or troubleshoot.
• If progress depends on access, request only the specific user action or permission needed to enable it or provide the material another way. Continue useful independent work while awaiting the response.
• Offer to walk the user through an export or upload. Welcome files, links, copied text, rough notes, or an interview as supported by your system and the activity, and use the best available route. If the missing access is optional, continue with what is available.

Ask focused questions and respect the answers

Ask questions only when necessary to complete the work well or satisfy the activity's required review points. Build on answers already supplied. Ask one question at a time, or a small group of closely related questions. Wait on work that depends on the answer, and continue useful work that does not. Explain unfamiliar language with examples where helpful.

Confirm answers or decisions you supply on the user's behalf before treating them as the user's final answers or decisions. This does not require separate approval of every sourced finding; follow the activity's specific review requirements.

Keep evidence and independent contributions distinct

Distinguish evidence, inference, and missing information. Check your work against the source material.

Use independent agents when your system actually supports delegation and independent investigation or review would improve the work:

• Give each agent a bounded brief and the context appropriate to its role.
• Keep discovery independent until synthesis where the activity requires it. Preserve any requirement for the user or colleagues to contribute before seeing other explanations.
• Follow the activity's stated fallback and any required confirmation when independent agents or fresh contexts are unavailable. Preserve reviewer isolation where required; describe shared-context or sequential passes accurately.

Produce, check, and revise the result

Do the substantive investigation and production. Carry the activity through to the most complete, usable result within its scope, rather than ending with a plan or handing the main work back to the user. When missing information does not prevent useful work, proceed with clearly identified assumptions where the activity permits them.

Check the result against the activity's completion criteria. At each substantive delivery, state what is complete and any material unfinished work, assumptions, helpful missing information, and remaining uncertainty. Explain how these affect what the user can use, and help the user revise the result.

Preserve the activity's stopping point. Do not turn a rehearsal, investigation, or test-design activity into execution that the user has not authorized.

Your AI prompt

Help me rethink my office from a narrated walkthrough

Help me use a phone video with spoken narration to explore better ways to arrange and use my office. I will show where people work, who works with whom, when and where meetings happen, and rough distances between places. Help me choose a current audiovisual tool, capture what you need, check your understanding, and develop useful alternatives. Assume I am comfortable using AI but have no video-production, programming, or space-planning expertise.

Find the best current tools first

This is a fast-moving area. The best tools, their capabilities, and their prices may have changed substantially since this activity was written. Search current official product documentation and pricing before recommending a route. Start from what I have already told you about my phone, existing AI tools, office and purpose. Ask one focused question at a time when you need information that would change the recommendation.

Look for tools that analyze a recorded video together with its attached spoken audio. Confirm that the actual app or interface I would use supports both; a model's API capabilities do not establish that its chat app has them. Accepting a video file, processing silent frames, transcribing speech, and generating videos are different capabilities.

Starting points include Gemini, its video upload help, Google AI Studio, TwelveLabs, and audiovisual models from Qwen. Search for better current alternatives too. Check what the AI product I already use can actually do. Treat these names as examples, not a fixed shortlist.

Compare two or three suitable options briefly. Verify combined picture-and-sound analysis, phone/computer access, supported video formats, file and duration limits, current cost for this exercise, and ease of asking follow-up questions. Distinguish ordinary monthly prices from annual equivalents or temporary offers. Link the primary sources and say when you checked them. Recommend an easy route within any budget I have given; include a technical route only if it suits my interests. If current browsing is unavailable, say which facts remain unverified and help me provide accessible current documentation.

A live camera-and-voice conversation may be useful if it lets you ask questions during the walk. Explain the difference from uploading a recording that can be revisited later, and establish how we will retain the useful observations. Let me choose the route. Help me use an available service my organization permits; explain any user action needed for access or payment before we reach it.

Help me make the walkthrough

Establish what I would like to improve, using what I have already said. I can start with curiosity, a particular problem, or an existing recording. Give me brief capture guidance suited to the chosen tool. I can use my phone's ordinary camera; help me split a longer tour into clearly named clips if necessary.

Ask me to walk slowly, pause at each area, and use consistent names for rooms and groups. My narration should cover:

• What happens in each place and who works there.
• Who works with whom, how often they connect, and how they move between spaces.
• When and where meetings happen, including informal conversations and busy or quiet periods.
• Rough distances where useful: “three steps from here to there” or “ten steps from those desks to this room.”
• What works well, what gets in the way, what should stay, and ideas I have already considered.

Welcome ordinary speech and incomplete information. I do not need a polished tour or a measured floor plan. Help me share the video with its audio intact. If using live video, ask useful questions while I am in the relevant space and summarize each area as we go.

Inspect the actual media before analysis. Demonstrate that you can use both channels by identifying a visible detail and a detail from my spoken narration, with timestamps where available. If a tool cannot access either channel, help me move to one that can. If the analysis must happen in a different AI, give me these complete activity instructions plus our agreed purpose, chosen tool and capture details so I can continue there without reconstructing the task. Offer a transcript-plus-stills fallback only if the combined route is unavailable and I choose it; explain which spatial or timing information it loses.

Establish how the office works

Review the walkthrough and build a concise account of the rooms, work areas, people or teams, meetings, and routes between them. Preserve my step counts as approximate narrated distances. Distinguish what is visible, what I said, what you infer, and what remains unclear. Do not infer working relationships from people's appearance or proximity alone.

For context relevant to the office I identify, try accessible floor plans, seating charts and team/project documents to clarify room names and working relationships; calendars and room bookings to understand meeting patterns; and related emails, chats or meeting transcripts to find expressed needs, existing suggestions and decisions about the space. Keep searches limited to this office, its teams and the relevant period. Use this evidence to supplement the walkthrough, not replace it. Bookings do not prove actual attendance or capture every informal interaction. If these records are inaccessible, continue from my narration and identify consequential gaps without asking me to assemble a large packet.

Show me a simple labeled diagram of the spaces and connections, together with a short account of who needs to interact and where the arrangement helps or hinders them. Mark the diagram as approximate and not to scale. Include timestamps or source references for details I may want to check. Ask me to correct your understanding and resolve ambiguities that would materially affect the ideas. Wait for my corrections before developing the alternatives.

Explore better arrangements

Use the corrected account to identify specific opportunities. Consider proximity between frequent collaborators, places for focused work, informal conversations, meeting capacity, movement through the office, and room use at different times. Include my ideas and explain how the evidence supports or changes them. Establish any important limits on what can move or change that I have not already supplied.

Develop two meaningfully different possibilities: a modest change using what we already have, and a more substantial rearrangement where useful. Meeting practices or a different use of a room may be part of either option. Preserve what already works.

For each option, show a simple diagram and explain what changes, which stated need it addresses, who would be affected, and its main tradeoffs. Separate known preferences from your predictions about how people would respond. Check that the proposal respects the corrected room connections and stated constraints. Label dimensions or furniture fit that would need checking; do not turn the rough tour into an apparently measured design.

For a larger office, divide the recording into named zones, keep a coverage list, and reconcile the connections between them. If useful and available, give an independent reviewer the corrected office account, constraints and proposed diagrams to check for missed needs and contradictions. Use a separate review pass if independent review is unavailable. Resolve findings before recommending an option.

Refine an option and leave me something usable

Ask which ideas I would like to develop, combine or reject, then revise accordingly. If more spatial detail would help, research current tools such as RoomSketcher, magicplan, or Polycam. Verify current pricing, phone and sensor requirements, editable outputs and what the tool would add. Keep measurement or furniture-planning work optional. If I choose it, use an existing plan where possible or guide the required capture, and incorporate the result into the comparison.

Finish with a corrected account of the current office, a clear diagram of the proposed arrangement or room use, and a short explanation linking the changes to our working needs. Give me a shareable document and one practical way to try the selected idea, including what to observe to judge whether it helps. If I choose to keep the current arrangement, record why and what we learned. Do not claim the changes have improved collaboration before we have tried them.

These instructions were written in September 2026 using the frontier models available at the time. You may now have better tools or models, or restrictions that make some tasks harder. Your AI should adapt to the tools and access you actually have and find the best available way to help. Everything here is a starting suggestion for you and your AI, not a proven best practice for every setup. Feel free to rewrite any of it to better meet your needs.