Understanding about Artificial Intelligence and how it works

Introduction

AI has been incorporated into our lives today. Most people would think of AI as Ghat GPT, Gemini, or Claude. In reality, these are just a part of AI. Understanding what AI (Artificial Intelligence) and how it works will enable us to use its tools better in the future.

In simple terms, AI is a machine model that learns things through machine learning and perform tasks based on what it learns. Think of it as a student who study for an exam. The student consumes knowledge through books and lecture notes. He then use that knowledge to take exams. AI works in a similar way. It consumes data as input and use it for various purposes. Take Chat GPT for example. It takes in data and use it to perform tasks we ask it to do.

Earlier, I said that AI learns things through the process of machine learning. Machine Learning is a process where we feed data as input and train the AI model to perform specific tasks. This can be done in three ways.

Supervised Learning - The AI model is trained to provide specific outputs
Unsupervised Learning - The AI model is trained to find patterns within the data
Reinforcement Learning - The AI is trained through trial and error, guided by the process of rewards and penalties.

Most AI that we are using today is called Large Language Models. They are models that trained under large amounts of data to find patterns and generate a response. For example, if the text data used to train is

My favorite food is cheeseburger

"My"

"My favorite food is cheeseburger"

In reality, a lot of data are used to train to model so that it can respond to the user in human language.
Chat GPT and Gemini are considered as a weak AI, a model that is trained to perform specific tasks but is unable to do anything outside it. It also lacks human conscience to make human judgements.

The key in using AI lies in prompting or giving instructions effectively so that it can execute tasks as desired. This includes asking good specific questions and approaching promptings through different techniques.

What we're using today is known as weak AI. In future, it is possible that with enough data, strong AI or even super AI will be developed

Tulayapornchoti P. (2026 April, 9), AI – The most valuable asset at disposal, Quiet Horizon, https://myquiethorizon.com/2026/03/31/artificialintelligence/