Condition Matrix, Privacy and accessibility checklist, Model Card
Open mission materials →
Concept art made with Gemini 3.1 Flash Image← Mission map
Mission 11 of 12Release audit50 to 65 minutes
Rainbow Rescue Festival
The festival must work in sun, shade, clutter, quiet, and noise, with a safe way for every guest to play.
Today's AI power
Responsible AI game design
Responsible AI uses needed inputs, shows limits, offers another way to play, and lets a person decide.
Start here
Clear the condition matrix, alternative-control route, privacy promise, and two human overrides.
The condition matrix is complete.
Vela joins the team and the Release Model Card
See how this game works
Game loop preview
See the power before you build it.
- 1SignalA player signal in a new condition
- 2AI moveThe result meets your release rules or stops
- 3Your ruleThe exact game logic checks what happens next.
- 4Game stateHUMAN DECIDES
Your mission
One step at a time. One AI power to unlock.
Step 1 of 9Story
Story
Meet the mission problem.
Story signal: The festival must work in sun, shade, clutter, quiet, and noise, with a safe way for every guest to play.
Do this now
- 1LOOK
Spot the problem in the opening story and tell your crew what needs to change.
Goal
Know what winning looks like.
Your choice: Choose when the system must stop for HUMAN DECIDES.
Watch this change: A failed condition pauses the game and opens the accessible human route.
Do this now
- 2CHOOSE
Win by proving this: The condition matrix is complete.
Get ready
Gather the props and get your grown-up guide.
Do this now
- 3CHECK
Open the Mission 11 materials on screen or print them, then gather the listed props.
- 4CHECK
Classify props and game signals only. Use HUMAN DECIDES for every question about a person or human trait.
Predict
Guess first, then test.
Do this now
- 5CHOOSE
Run the prop game across the condition matrix.
Build
Make the AI power in the Google tool.
Do this now
- 6MAKE
Follow the Google tool card below. Complete each step in order, then bring the visible result back here.
Google tool step
Open the tool, make the power, then bring one result back.
OPEN TOGETHER
Teachable Machine
Adult and child operate together
Build thisExisting prop classifier preview
Save asRelease Model Card v1
Do these steps
- Open the prop classifier preview.
- Test each listed non-person condition.
- Record pass, limit, or stop.
- Demonstrate the alternative control.
- Stop the tool when HUMAN DECIDES is required.
- Stop camera or microphone access after testing.
You are done whenThe matrix shows where the model works, where it is limited, and where a human route takes over.
Need setup help?
- Bring
- Prop-only model
- Condition Matrix
- Privacy and accessibility checklist
- Alternative control
- Next move
- Complete the Model Card and make a manual release or stop decision.
- If stuck
- Use recorded sample outputs and the alternative control if the external project cannot be opened.
Bring it back
Return with the result you can see.
Do this now
- 7CHECK
Complete the model card with purpose, limits, tests, and override. Save the exact mission artifact as Release Model Card v1, then return to this card.
Glitch test
Make a funny mistake on purpose.
Funny glitch: A noisy booth asks for help and Mochi answers by pressing the giant touch button with both paws.
Do this now
- 8TEST
Move the game into a busy or noisy scene and watch a shortcut or accessibility gap appear.
Boss round
Change one thing and retry.
Game move: Multi-booth rescue festival with the full cast
Harder remix: Defend the privacy boundary, accessibility alternatives, and two override cases in a short release review.
Do this now
- 9REMIX
Choose one evidence-based change, run the same test again, and compare what changed.
- 10TEST
Run the boss round: Two HUMAN DECIDES examples are recorded.
Explain
Say what the AI did and prove it.
Power reward: Human Override Captain · Vela joins the team and the Release Model Card
Responsible AI uses needed inputs, shows limits, offers another way to play, and lets a person decide.
A model that works only for one room or one way of playing can exclude people or make unsafe choices.
Do this now
- 11SAY
What makes an AI game responsible beyond simply getting more correct answers?
- 12SAVE
Complete the mission proof and save only the named artifact: Release Model Card v1.
What counts as finished?
What makes an AI game responsible beyond simply getting more correct answers?
- The condition matrix is complete.
- The privacy promise bans person and sensitive-trait inference.
- At least one alternative control reaches the same goal.
- Two HUMAN DECIDES examples are recorded.
Need help or want a harder challenge?
Make this step easier
Use picture stations and a large coral human-choice card with sentence starters.
Say it this way: The AI noticed ___. The exact game rule ___. I know because ___.
Try the deeper remix
Defend the privacy boundary, accessibility alternatives, and two override cases in a short release review.
- Record the inputs and outputs.
- Compare two versions or conditions.
- Name the constraint or exact rule.
- Design one fair remix test.
Everyone proves the same AI idea: The condition matrix is complete. · The privacy promise bans person and sensitive-trait inference. · At least one alternative control reaches the same goal. · Two HUMAN DECIDES examples are recorded.
Save points for this longer mission
Save points
Pause here. Resume without losing the plan.
Use these on-device cards to remember where you stopped and what to do next.
Quick power check