Page 27 - Ai Robogenius

Now, let’s explore another type of learning.
Instead of explicitly programming rules such as "if this happens, then do that", we let the system

discover patterns or rules by itself that help solve the problem.
How do humans make decisions?
For example, imagine we are preparing a glass of lemonade, but we don’t know exactly how

much sugar to add.
Instead of following a strict rule, we taste the lemonade after adding some sugar, then adjust the
amount based on how sweet it is. Over time, we learn the best amount of sugar for our taste by

trying, observing and adapting.
This process of learning from experience rather than following fixed rules is similar to how some
AI systems learn and improve.

Time of lemonade making Amount of sugar added Taste of lemonade
First time Very little sugar (quarter teaspoon) Not sweet enough
Second time Too much sugar (two teaspoons) Too sweet

Third time Moderate sugar (one teaspoon) Just the right sweetness
Fourth time Moderate sugar (one teaspoon) Just the right sweetness

21 st
Century #Media Literacy
Skills
Video Session

Watch the video of "How does ARTIFICIAL INTELLIGENCE work?" at the
given link:
https://www.youtube.com/watch?v=fLcZRDiQyNY OR scan the QR code
and answer the following questions:

1. How do characters in the video decide which direction to take?

2. What do the different colours in the video represent?

3. How does Rule-Based AI use predefined rules to reach a destination?

Learning-Based AI
A learning-based approach is when a machine learns by itself from the data it receives, instead

of following fixed rules made by a person. When the data is unlabelled or random, the machine
looks for similarities and groups similar data together.

Stages of AI Project Cycle
25