Welcome to the fifth lesson! You will learn:
- How to use AI to change your community in different ways
Materials you will need:
- AI Inventors Notebook
- Brainstorming materials: You can use an online brainstorming tool like this or paper and markers
Key Slide Notes:
- Motivation video about Machine Learning (15).
- Not every problem requires technology or AI (23).
- The more data, the more patterns a model can find. Data helps AI agents understand the world. AI without data won’t achieve its goals well (32).
- Your invention will use data (audio, images, text, numbers) to make decisions and take actions towards a goal (34).
- An AI agent more data will achieve its goals better than an AI agent with less data (34).
Activity #1: Lightning Brainstorm
For every question or picture, think of an answer. Try to think of the most creative thing you can say. The more unique the better.
- If you had 2 minutes with your political leader. What would you tell them to improve about this place?
- Think of one activity you want to do faster or better. How would you do that faster or better?
- How would you make a school classroom environment more fun?
- Are there health systems in your community that could be improved?
- Imagine you are a superhero… who or what on our planet would you help?
- What is one thing that exists in this place you’d like to make better?
- What is a problem in these places or something you’d like to work better?
- Pick 2 of the problems you’ve discussed that matter to you.
- Talk about solutions that might solve the problems
Activity #2: Make a community map
You can use papers and markers or an online tool like this to do this activity.
- Pick one idea to continue developing
- Create a Community Map. Write your community’s name in one box (this will be your main one). Then add these four sub-boxes: before solution, after solution, invention, AI.
- For the before solution box, what is your community like before your solution? Questions you should ask: 1. What does it sound and feel like? 2. Who is impacted? 3. What does the problem look like? What’s broken or needs to be fixed?
- What is your community like after your solution? Questions you should ask: 1. How does your solution change things? 2. What’s better for people? 3. What looks, works, or feels different with your solution?
- Think more about your invention by answering the Invention questions. Questions you should ask: 1. How would a human “expert” act to solve your problem today? 2. Whose life will be better if you solve this problem? 3. How will your invention solve this problem? 4. What data does your invention need to solve this problem?
- Finally, consider these questions about AI. Questions you should ask: 1. What actions will your invention take with new decisions? 2. What decisions will your invention make to ensure the situation improves? 3. What are at least 2 patterns it will find in the data? 4. How could your AI agent gather data?
Community Map Example:
- Community name: NYC
- Before solution: Safe bike lanes are unknown. 1. What does it sound and feel like? Busy streets, blocks bike lanes. 2. Who is impacted? Frustrated bikers. 3. What does the problem look like? What’s broken or needs to be fixed? People on bikes.
- After solution: 1. How does your solution change things? More people on bikes. 2. What’s better for people? Safer environment for bikers. 3. What looks, works, or feels different with your solution? Bikers only use open routes, can avoid the blocked ones.
- Invention: 1. How would a human “expert” act to solve your problem today? An expert would check every bike lane, all the time and update which ones are closed. 2. Whose life will be better if you solve this problem? Bikers’ lives would be better. 3. How will your invention solve this problem? My invention will be constantly updating which lanes are closed. 4. What data does your invention need to solve this problem? Needs data on all bike lanes in the city, and when those lanes are closed or open.
- AI: 1. What actions will your invention take with new decisions? Change the routes. 2. What decisions will your invention make to ensure the situation improves? Recommendations on the route. 3. What are at least 2 patterns it will find in the data? Which routes are unpredictable and how long blockages last. 4. How could your AI agent gather data? Riding time for cyclists and view of the streets.
Activity #3: Compare these 2 inventions:
Which of these two inventions needs AI?
Data the inventions gathers:
- Images of people in pools.
- Measurements of angles.
Does the inventions find patterns in data to make decisions?
- Yes – the more data about swimmer’s motions the better it is at recognizing if a person is drowning.
- No – there are rules about angle measurements to follow.
What decision does the inventions make?
- Decides if swimmer is drowning or okay.
- Decides if the angle of the line is acute, right or obtuse.
Can the invention take new actions based on different decisions?
- Yes. It can alert if people are drowning or not based on different decisions about motions
- No. No matter how much data it gathers, there won’t be different decisions.
Activity #4: Research your idea
Research and complete your idea in your AI Inventor Notebook:
- Find statistics about the problem online. Are there multiple places this problem occurs. How many people are impacted? Are there other organizations or companies addressing the problem you can learn from?
- Create a survey to collect information from people in your community
- Find an expert you’d like to interview. Create a diagram of what you know and what you want to know. Then, write out the questions you want to ask them.
- Identify 2 competitors and think about how your idea is different
Activity #5: Which AI agents works better?
- 1) An AI agent that has 1000 hours of your watch history data and recommends you videos it thinks you’ll watch. 2) An AI agent that recommends you videos based off of 200 hours of your watching history. Answer: 1) 1000 hours of watching history is better
- 1) A robot that picks up cups that has practiced with a coffee mug. 2) A robot that picks up cups that has practiced with a coffee mug, a glass jar, and a teacup. Answer: 2) Robot practicing on more cups has learned from more data.
- 1) Cake decorator with 20,000 pictures of beautiful cakes. 2) Cake decorator with 30 pictures of beautiful cakes. Answer: 2) Cake Decorator with 20,000 pictures.
Activity #6: How could you make this dataset better?
- This dataset could use different colors of cars/planes, different models/types, different textures, backgrounds, etc.
Activity #6: What patterns in the text data might make an AI agent incorrectly match the statements to the right picture?
Text data:
- Cold blooded animals
- Eat leaves from tall trees and drinks water
- Lives in water
- The tallest living terrestrial animals
- Uses water in emergencies to put out fires in tall buildings
- Maintains law and responds to emergencies
- Emergency rescue for animals in tall trees
Water is mentioned for multiple, animals mentioned for multiple, rescue mentioned for multiple, tall
Reflection:
Review the options and pick which invention would make the best use of AI:
- An agent that tells you the steps for assembling a toy car
- An agent that gathers people’s previous shopping lists and predicts what they need to buy on their next trip
- An agent that gathers sound data for a database of free sounds
Answer: 1) This agent is making decisions and taking action to tell a person what to buy on their next shopping trip.