With the launch of Gemini 2.0, a powerful new language model, Google has taken its AI to the next level. This update brings some impressive upgrades, making it even better at:

• Text Generation: It can create human-like text, translate languages, write all sorts of creative content, and answer your questions with useful information.
• Code Generation: It can write and fix code in multiple programming languages.
• Reasoning and Problem-Solving: It’s great at handling complex tasks that require logical thinking and problem-solving.

Gemini 2.0 Flash is built for speed and is perfect for real-time chat, interactive assistants, and quickly pulling up information when needed.

The purpose of this tutorial is to introduce you to Gemini 2.0. We will explore the settings and gems. Gems are like custom GPTs. You get a pre-coded chatbot that answers specific prompts like coding assistance, a career guide, or a brainstorming assistant. After exploring the settings and the gems, we will show you some real-world examples where Gemini 2.0 shines by assisting you in specific tasks like generating images or finding solutions to coding issues. 

In this tutorial, we will cover:

• Gemini 2.0 overview and features
• A simple real-world example with Gemini 2.0

You’ll need:

• Gemini Advanced (Free Trial)
• Google AI Studio

Gemini 2.0 overview and features

Alright, so let's dive into how Gemini 2.0 works under the hood. It's designed with a modular architecture. Each module focuses on a specific task, and they’re all connected to build a smooth flow of data and processes.

‍Core Modules

First up, we've got the core modules. These are the main building blocks of Gemini 2.0, and each one handles a key part of the workflow:

Data Ingestion Module: This is where it all begins. The ingestion module pulls in data from a variety of sources. It could be data from your company's database, an API you’re connected to, or even from files like CSV or JSON. Think of it as the “data catcher” that brings everything in.

Data Preprocessing Module: Once your data is in, it needs some TLC. This module helps clean up messy data, handle missing values, remove duplicates, and normalize it. It’s all about getting your data into shape so it can be used effectively in your analysis or machine learning models.

‍Model Training Module: Now we get to the fun part—training your machine learning models. This module is where you feed the preprocessed data into an algorithm and train it to recognize patterns. You can either use one of Gemini’s built-in algorithms or customize it to fit your specific needs.

‍Model Evaluation Module: After training your model, you need to check its performance. This module evaluates accuracy, precision, and recall. It tells you if the model is working as expected or if you need to adjust it to improve performance.

‍Deployment Module: Finally, once you’ve got a model that’s performing well, you need to deploy it. This module lets you deploy the model into production, either as an API or integrated into another system. It’s all about taking your model from the development environment and putting it to work in the real world.

Important to know: There are a couple of ways to access Gemini 2.0. Go to Gemini 2.0. A ChatGPT-style chatbot will greet you. If you want to access Gems and other settings, you have to subscribe to Gemini Advanced. It’s free for the first month!

‍

‍

After starting the trial, you will see the usual AI chatbot page layout. 

‍

‍

Click Gemini Advanced at the top left corner of the chat window and select ‘Gemini 2.0 Flash’ (experimental).

‍

‍

You’re all set now. Start asking questions and see how it goes. 

Another way to access it is by visiting Google AI Studio. You can access Gemini 2.0 for free. It’s limited but you can check it out. However, if you have subscribed to Gemini Advanced, you can see all the settings and features. 

‍

‍

Google Gemini is better for beginners and casual users. AI Studio is for serious enthusiasts and professionals who want to use it with their development codes and functions. 

Let’s focus on the Google Gemini chatbot first. We will go to AI Studio in the integration and preprocessing topic. 

Gemini 2.0 is very much similar to OpenAI’s ChatGPT. You can ask it to create a table of daily tasks, or upload an image and ask it to modify the image. You can also use voice input to ask a question.

Prompt:

Create a table of {a brief description}

‍

‍

That’s really an exhaustive list. Not bad considering that Gemini 1.5 was not that useful. Let’s see what the image uploader can do. Click the image button at the left of the input box. 

‍

‍

Upload an image and write a prompt. In our example, we have uploaded an image of a 1969 Porsche car. 

Prompt:

[Upload image]
Transform the Porsche 911 into a roofless car.

‍

‍

It generated an image without the roof. 

‍

‍

Experiment with your prompts and see what it can do. Let’s see if it can generate an image on its own. 

Important to know: Gemini 2.0 Flash is still under development. It cannot generate images of people yet. 

Prompt:

Create an image of a flower pot with a growing cactus plant.

‍

‍

Another cool feature is ‘Listen’. You can listen to the answer by clicking the three-vertical-dot icon at the bottom of the generated answer. 

‍

‍

There’s another feature you can use when writing prompts. Click the microphone icon on the right side of the text input box and ask a question. 

Prompt:

Can you give me a list of the rarest plants in the world?

‍

‍

The microphone doesn’t work well. I had to repeat the prompt multiple times before it figured out what I was saying. I was using an American accent!

Other features include settings, activity in Gemini apps, and Gems Manager. We’ll get to the Gems manager in a moment. First, let’s see which settings we can configure to make it more useful. 

‍The settings link is in the sidebar. Hover over your mouse to view an expanded menu. 

‍

‍

You can manage your subscription, change the theme, and access your public links and Extensions. You can also view recorded information. The recorded information takes you to Google’s settings page. Here you can enable Google to record activity in Gemini Apps. There’s an option to delete the activity as well. 

‍

‍

The Extensions link takes you to a Google Apps page. You can enable Gemini in Google apps such as Workspace, Google Hotels, Maps, and YouTube. Slide the switch next to each app enabling Gemini to access it and provide suggestions. 

‍

‍

Your public links let you share discussions or specific questions and their answers. You can easily manage and view all the public links you've created right here. It’s related to Google Workspace smart features and personalization settings. 

‍

‍

Gems

Gems are like custom GPTs.There are four predefined gems offered by Google. 

• Brainstorming assistant
• Career Guide
• Code partner
• Coach to learn

Brainstorming assistant

It’s a basic gem. It helps you brainstorm ideas. For example, if you are planning a party and you lack inspiration, It can suggest original themes, unusual decorations, fun activities, and even tasty recipes.

It can also help you find gift ideas for your loved ones, taking into account their tastes and your desires.

‍

‍

Prompt:

Give me ideas on {a description of whatever you are thinking}

‍

‍

Career Guide

It’s a unique custom GPT-type module that acts as a personal career coach, designed to help you advance your career. If you’re looking to improve your skills, achieve new goals, or navigate the complexities of today’s job market, “Career Guide” can help you with pro tips and tricks.

According to Gemini, a career guide can create a:

• A personalised plan: Your unique needs and goals are at the heart of the approach.
• Expert advice: Benefit from the expertise of seasoned professionals in the field of career coaching.
• Tools and Resources: Access a variety of resources to help you succeed.
• Ongoing support: We are here to encourage and motivate you throughout your journey.

Let’s see what it can do. Start asking questions. Use the following prompt:

Prompt:

How to write a convincing proposal for a job {specify a job role and a brief description of what you do and your experiences}

‍

‍

You can ask career-related questions. A few example prompts you can use in the career guide are:

• How can I improve my presentation skills?
• Help me understand how to advocate to my superior for a promotion.
• Give me some advice on how to find a mentor.

Code partner

Another great gem is a code partner. The name says it all. It helps you code. A code partner gem is a Ruby gem designed to assist developers with their coding tasks. It acts as a collaborative tool, offering features like code completion, real-time feedback, and debugging assistance.

Think of it as a tireless pair programmer that can help you write better code, faster.

Let’s find out ourselves. I am going to paste an incomplete Python code and ask it to complete it. 

Prompt:

Can you complete this code for making an API call:

import requests
def get_posts():
   url = 'https://jsonplaceholder.typicode.com/posts'
    try:
        response = requests.get(url)
        if response.status_code == 200:
            posts = response.json()
            return posts
        else:
            print('Error:', response.status_code)
            return None
    except requests.exceptions.RequestException as e:
        print('Error:', e)
        return None
def main():
    posts = get_posts()
    if posts:

‍

‍

It completed the code perfectly. We ran it and it executed flawlessly. Gemini 2.0 Flash is an amazing tool for coders/programmers. It can figure out problems in code and return better results. 

Gem Manager

Gem manager helps you manage Ruby gems. It simplifies installing, updating, and removing gems from your projects. Think of it like an app store specifically for Ruby code libraries.

Popular examples include RubyGems, Bundler, and Gem. They handle dependencies, ensuring you have the correct gem versions, and keep your projects organized.

You can add your gems in Gem Manager. Start by clicking ‘Gem manager’ in the sidebar.

‍

‍

It lists predefined Gems added by Google. These are the default ones that you can access anytime. 

‍

‍

You can create a new gem (custom GPTs) by clicking the ‘New Gem’ button in the Gem manager. We will focus on creating a ‘New Gem’ in another tutorial. 

A simple real-world example with Gemini 2.0

Now comes the fun part. We will try a real-world use case with Gemini 2.0 to see how far it can go. For this example, go to Google AI Studio. Click ‘Stream Realtime in the sidebar.

‍

‍

Important to know: We used Stream Realtime because Gemini 2.0 couldn’t connect to the internet to get the phone numbers and addresses of the restaurants when we used the prompt in the “Create Prompt” window. It looks like Stream Realtime can access the internet, while the other modules can’t.

Let’s find out all the Italian restaurants in New York City, their phone number, addresses, and brief descriptions, as a table. Copy the following prompt:

Prompt:

 I want to find these Italian restaurants in NYC. Please provide a phone number, address, and a brief description as a table: I want to look up all the Italian restaurants in NYC. Please provide a phone number, address, and a brief description as a table: Lucciola, Carmine's, via Carota, and Osteria Barocca.

‍

That’s a very good response. The notes are informative and to the point. Gemini 2.0 is better than the previous versions. It can definitely be used as a coding and Live API tool to save time and solve petty issues. 

‍