Page 51 - CT_AI_Class-8

ethical minds

AI systems in healthcare and environmental monitoring may reinforce biases if trained on
unrepresentative data, leading to inaccurate outcomes. Ensuring diverse and inclusive datasets is
crucial to provide fair and accurate predictions for all demographics and regions.

NO-CODE AI

No-Code AI is a way of building artificial intelligence (AI) systems without needing to write any

computer code. Normally, creating AI systems requires programming skills, but with no-code
tools, you can create AI by simply dragging and dropping different features or blocks, much like
putting pieces of a puzzle together.
These tools have easy-to-use interfaces that let you design AI systems to do things like recognise
pictures, analyse data or even chat with people, all without having to understand complicated
coding languages.

No-Code AI is great for learning and experimenting with AI in a fun and simple way.

Popular No-Code AI Platforms
 Teachable Machine (teachablemachine.withgoogle.com)

 Machine Learning for Kids (machinelearningforkids.co.uk)
Both platforms provide step-by-step tutorials and video guides to help you get started.

Hands-On Activity 2: Voice Command Recognition

Problem: Create an AI model that recognises three voice commands:
 Start

 Stop

 Help

Steps to Follow:

1. Set Up Your Project

 In Teachable Machine, start an "Audio Project"

 Create three classes: "Start," "Stop," and "Help"

 Add a fourth class called "Background Noise"
2. Record Training Samples

 For each command, record yourself saying it 15-20 times

 Vary your voice—say it loudly, softly, quickly and slowly

 For "Background Noise," record 20 seconds of silence and room sounds

Deeper Dive into AI Applications 49