Getting Started With Google Gemini

Introduction

With AI rampaging around the world nowadays, it is essential to learn about it. One of the places to start is Large Language Models (LLM). Think of Large Language model as an intelligent chat bot that can answer anything and everything you want it to answer provided that you can control them well. Google Gemini is an example of LLM

LLM is trained under Machine Learning where it learns from a set of large data in order to try to answer our questions. If a question is asked, it will construct its answers through predicting words in a sentence iteratively. It is no exaggeration that Large Language models such as Chat GPT and Google Gemini acts as search engines tailored to our requirements today, provided that it is trained well. While most people use it as search engines, more capable users can use it more efficiently.

Fast mode allows the model to answer questions instantaneously without thinking. The answers may not be detailed and is used for solving simple problems or questions such as

Summarize the following context for me in 3 key points

Thinking mode makes the model perform full analysis on the questions and usually is used for solving more complex problems that required finer details.

Deep Research is like Google Gemini in thinking mode, it takes longer. Think of Deep Research as a university graduated student who is working on final project. It takes much longer, but the answers are more accurate and refined.

Canvas Mode is using for writing and editing post in the process. Say you want the model to write the article, The model writes it. You want to edit one section of its answer. Canvas Mode allows you to do that.

Personal Context basically tells the model who you are as a person. Tell the model about yourself there. What you like? What are you currently doing? How would you like to communicate with the model? The answers generated by the model will be tailored to your specifications based on your personal context.

GEMS are like mini LLM models create for specific purposes. One may be a food expert in Bangkok. Another may be a writing expert acting as William Zinsser. Instead of a regular search engine, it will perform tasks stated in requirements you put in GEM.

Satangmongkol K. (2026), Mini Gemini Bootcamp, The School of Duck, https://www.duckstep.co/mini-gemini-bootcamp
Tulayapornchoti P. (2026 April, 9), AI – The most valuable asset at disposal, Quiet Horizon, https://myquiethorizon.com/2026/03/31/artificialintelligence/