In this ultimate Gemini tutorial for
beginners, I will show you really EVERYTHING about Gemini – and by "everything"
I mean really everything! From the basics, the latest features and updates of Gemini,
to prompt engineering and much more. I will illuminate every single topic for you step by
step in this video, so that by the end of the video you are a real Gemini pro. Oh yes, and as a
small thank you for watching the complete video, there will be a special document from me
at the end, in which I have summarized everything that is important for you again!
With that, let's get started directly. First and foremost, you naturally need to sign
up for Gemini, and you do that with your Google Account. Just click on 'Sign in' and enter
your details. However, if you have signed up but can't use Gemini because your account is not
supported, it's probably because your age needs to be verified. In such a case, you just need to
send Google an ID or driver's license to confirm your age. You can find all the necessary
links for this in the video description. Once you've successfully signed up for Gemini and
qualified, you'll land on the Gemini interface. In this video, I'll be using Gemini Advanced to
demonstrate the features of Gemini, which is why I'll first select Gemini Advanced from the menu
at the top, where you can choose your AI model. You might be wondering: What's the difference
between Gemini Advanced and the regular Gemini? Simply put, Gemini Advanced is Google's new AI
model that offers improved features and operates more efficiently than the standard Gemini model.
While Gemini is available as a free version, Gemini Advanced costs 20 dollars or approximately
22 euros per month in Europe. My tip: You can try Gemini Advanced free for two months, which I
highly recommend. At the video's end, you'll get a document outlining who should and shouldn't
buy the paid version, plus a model comparison. You might have already noticed that there are
four fields here. These are example prompts, which I won't be using because I'll be using
my own. But what exactly are prompts? A prompt is a message that you enter below and
then send to Gemini to receive specific information or an answer. Keep that in mind
for now; we'll come back to it later. First, I want to introduce you to some of
the features of Gemini that I often use. This will give you an impression of
what is possible with Gemini. Don't worry, we will get to the interface later. You should
first learn what you can do with Gemini. I'm starting now with creative writing.
Theoretically, I could ask Gemini to write a story or a poem. Instead, I'm asking
Gemini to create an email template for welcoming new employees, including all the
important information and a warm welcome. I choose this example because most will
likely use Gemini for writing emails. After entering the prompt and pressing enter,
you can see how quickly Gemini responds and creates a suitable email template. I
won't read the template out loud here, but you can stop the video and read it
yourself. What I can tell you, though, is that it's really well done and even
includes explanations for why certain elements were chosen. You see, even with this
first feature, Gemini can be a huge help, especially when it comes to emails. Now, let's
take a look at Gemini's multilingual capabilities. Since Gemini can write and understand many
languages, you can also have your texts translated. I will demonstrate this with
the translation of my email template using this prompt: 'Translate the email template
into Hindi, German, and French. Keep the tone and intent of the original, and ensure
correct grammar and syntax in each language.' As you can see, it now translates the template
into Hindi. I can't tell you if the translation into Hindi is perfect, but those of you who know
Hindi can write to me in the comments. However, I already see that it has not translated
everything into Hindi below. The German version seems to have been done quite well,
except for the lower part. Keep in mind that Gemini makes mistakes from time to time, so you
should always check the result again. In French, it could not generate everything for me, as
Gemini only generates certain text lengths depending on the AI model you use. Therefore, I
will enter another prompt to get the other part. It will then deliver the French version to me,
which I also cannot understand. Anyone who knows French can write to me in the comments if it is
a good translation. The lower part was also not translated there. You can, of course, write
another prompt to Gemini to have it translate this part as well. I won't do that now, as we
move on to the next feature: 'Idea Generation'. I'm now letting Gemini brainstorm 10 new ideas
for the welcome email to make it more informative and engaging. You can already see that some
great ideas are being given here. For example, I particularly like 'Interactive Elements'
and 'Resource Hub Link'. I mostly generate ideas for my YouTube videos and can really
recommend asking Gemini for a few ideas, as it's a good source of inspiration. Before we move on to the next features, I want to
clarify an important thing. You see that I have covered these three features in one chat because
they belong to one topic, "Email Template". However, if you have a prompt that has nothing to
do with the previous topic, it's better to switch to a new chat. That's because Gemini generates
your next message based on what you have discussed with it so far. In other words: It takes into
account the content you have already introduced. Of course, you can handle everything in one
chat, as Gemini can also flexibly switch between different topics without significantly affecting
the quality of the responses. But to ensure you get the best possible answer, it's better to open
a new chat so it focuses only on that one topic. I will do this now for the demonstration
of the next feature, "Grammar Check". For that, I go here to "New Chat", open a
new one, and enter the appropriate prompt, which reads: "Can you check
the grammar and spelling off this sentence: The dog walk down the rode
and eated his bone." As you can see, it corrects everything and even provides a suitable
explanation for why something was corrected. I often use this feature in Gemini
because I don't have the 40-message limit like in ChatGPT. But let's
move on to the feature "Summary". Gemini is very good at summarizing texts. I'll
demonstrate this with the following prompt: "Summarize this news article into
a 3-sentence overview. Focus on the main events and key information:..." I've
copied the news article below the prompt so that Gemini can access it. As you can see,
Gemini has concisely summarized everything in three sentences. This function is one of the
most helpful because it allows you to gather information more quickly. I will show you later
a better method for summarizing Newspaper. Now, let's move on to the feature
"Explaining Complex Topics". What's really practical about Gemini is that you
can have topics you don't understand explained to you simply. I'll demonstrate this with
the prompt here: "Explain Einstein's Theory of Relativity in a way a middle school student
would understand." Of course, I've already read the answer beforehand, and the explanation was
definitely very simple, so almost anyone could understand it. As mentioned, you can stop
the video and read the answers if you like. In this video, I will only read the most
important things so we can make progress. So, now I've shown you the basic
features of Gemini, and there are many more, such as image generation and
recognition. We'll get to that shortly, but first, I want to give you a
brief introduction to prompts. You already know that prompts are
messages you send to Gemini to receive specific answers. These prompts
can come in numerous variations, such as questions or tasks – the
possibilities are virtually limitless. It is important to understand that
the more specific you make the prompt, the better the AI model will deliver a
fitting answer. While this is mostly true, a longer prompt does not automatically mean you
will get exactly the answer you are looking for. Sometimes, a shorter formulation can
work better, depending on the task you give to Gemini. The key is that the prompt is
clear and contains only relevant information. The most common prompts you will use
are always connected with a Task and your Details and Context. For example,
a basic prompt could look like this: The Task: "Write a blog post." The
Details and Context: "About the latest developments in AI technology
and how they can change daily life." I call these two components a Basic Prompt,
the most frequently used elements of a prompt. The Task is the most important because
without it, Gemini would not know what it is supposed to solve for you. Thus, a Task
should always be included in your prompt. You add Details/Context to make the answer more
precise, to improve the quality of the response, and to ensure that the resulting answer
is as relevant and useful as possible. To clarify, here are three earlier
prompts with these two components. Feel free to pause the video
to take a closer look at them. Also good to know: Creating suitable
prompts to receive the best possible answers from an AI is known as Prompt
Engineering. And this involves more than just the two components I have just
introduced to you. Later in the video, I will introduce the other four components that
you can optionally incorporate into your prompt to achieve the best possible result. I call this
construct an Advanced Prompt. But first, I do not want to overcomplicate things; we will come
back to this later in the video, so stay tuned. Before we delve further into the
features of Gemini, I'd like to show you the chat functions that can make the
conversation with Gemini even more productive. One of the most sensible chat functions are
drafts. With these, you can generate multiple answers for your task. Conveniently, you can
also regenerate them whenever you want, which is very useful since you don't have to resend the
prompt each time and can directly compare answers. Another useful feature is that you can listen to the message from Gemini. I'll
demonstrate this to you right now. It's a really good feature to have while chatting. Below the message, you can modify your response.
For example, you can change the tone of the message – be it casual, professional, or simple
– depending on the task you have. You can also have Gemini write the answer shorter or longer,
which is especially useful if you're working on a blog or something similar. I'm going to generate
this message shorter to demonstrate it to you. As you can see, it has now become shorter.
Let's move on to the next chat function. Here, I've created a new chat with a
question I posed to Gemini to show you the next function. It's called "Double
Check," with which you can verify your answer from Gemini using Google. This
provides you with source references so you can check again yourself if the
question has been accurately answered. Below, you can share your chat – to Word,
Gmail, or just as a link, which can also be very handy. After looking at how to share the
chat, let's move on to another important function. With this function, you can edit your
prompt again if you want to make changes without sending a new prompt. This is very
sensible if you've made a mistake or want to improve it. As you can see, a new answer is
generated directly that matches my new prompt. Now, I'll show you what you can do on the
sidebar to the left. You can already see that your messages are displayed there, where you
can switch from one to another. When you navigate to a chat and click the three dots, you have
options to delete and rename the chat. Moreover, what I really appreciate is that you can
even pin your chat and assign an emoji to it, so you can always find the chat quickly. As
you can see here, I've pinned the chat with the Earth emoji, fitting the theme, and
thus can always find it quickly. Below, you can get help, view your chat activity,
and go into your settings. In the settings, you can manage your subscription, enable
real-time responses, change your theme, and manage your public links for the
chats you've shared. At the very top, we have extensions that we can activate. We'll get
to those in a moment. I'll show you first how you can better summarize articles or posts without
having to copy and paste the entire content. Since Gemini is part of Google, it can access
the internet and perform Google searches, which means you can simply copy a link from your
chosen article and add it directly to your prompt without copying the content. As you can see,
I've added it to my prompt here, and Gemini even recognized it and transformed the link into
the title. After sending the prompt, you can see that Gemini accesses the news through the internet
and responds with a short summary of the news. Now, let's move directly to
the extensions. Currently, there are only these five: Google
Flights, Google Hotels, Google Maps, Google Workspace, and YouTube. These
connect you with the Google Ecosystem, which is really handy. I will demonstrate
each one to you, starting with Google Flights. For this, I enter the prompt here: "Show me
flights to Madrid to visit my aunt during the week of February 22."
When searching for flights, the extension is triggered, and as you can
see, it directly gives me a response with recommendations. It even recognized which airport
is near me and directly recommended some flights, which I find really practical. Let me also
demonstrate this with the Google Hotels extension. My prompt for this is: "Find hotels for a 4-day
trip to San Francisco for New Year's Eve." Basically, the same principle
applies here as you can see it gives me recommendations regarding
hotels with descriptions and links. But now, let's move on to my favorite
extension. For this, I enter the prompt here: "Where can I go to eat in Frankfurt?".
This triggers the Google Maps extension, which I find really useful. It gave me some good
options for where I can eat. As you can see, there are many cool things possible with Gemini, but
this is nothing compared to what's still to come. you can access your email, drive,
and docs. I'll show you how this works. I'll start with the documents in my
drive. For this, I enter the prompt here: "Find my doc titled 'Weather in London?' from
my drive and summarize it." As you can see, this extension with Gemini accesses my drive and
my doc and summarizes it, which is really useful. You probably already have some ideas on how this
extension could make you more productive. Next, I give Gemini the prompt: "Are there any new
emails in my inbox?". This will check Gmail to see if I have any new ones. It looks like I don't
have any currently, but it shows me my last ones, which is also very cool. In my Third
prompt, I ask Gemini what's in my Drive. As you can see, it lists them and
even provides a description. Now, I'm coming to the last extension. With the YouTube extension, you can summarize
videos on YouTube, grab general information about them, and much more. I'll show you this
by taking one of my own videos and asking Gemini what it's about. You can already see that Gemini
has given me an answer, but it made a few errors regarding the topic. That's why it's important
that you always double-check the responses. Before I continue with the features, I will now conclude the topic of prompts so that
you really know everything important about it, o you can completely understand and generate
the best answer for your task with Gemini Since you're already familiar with the two
components of Task and Details/Context, I'll now unveil the other 4 components,
starting with tone and format. The tone determines the style in which the answer is
written, for example, professional, clear, casual, simple, as you've already seen with
the modify response in Gemini. With format, you decide the structure of your answer,
whether it's an email, bullet points, tables, or even Python code. Here's an
example prompt to better illustrate this: Task: Compose an summary of the last meeting Format: Email Details/Context: including
main topics, decisions/actions, and date/time of the next meeting Tone: in a clear and professional style. I believe through this example, you've understood
what Format and Tone are, as they are also the simplest components. Now let's move on to the
last two components, "Example" and "Persona". Examples are given to Gemini so it has a
template on how to solve your task more precisely according to your vision. An example
for the prompt we have would be this one. We've essentially told Gemini what
structure the email should be in, and it will then write the email that way
since we've added an example to our prompt. You should always add an example if you want to
tell Gemini exactly how to respond to something. A Persona, in short, can improve the effectiveness
and relevance of the generated answers. If you want your answer to be even more precise
and specific, it's advisable to add one. Here's an example of how I
added the Persona to my prompt: You are the team coordinator who
values precise and clear communication. By adding the Persona, Gemini will
solve my prompt even more precisely. So, these 4 additional components that you can
add to your prompt make what I call an Advanced Prompt. These are optional, and you can customize
the positioning and use according to your needs. If you didn't understand something about
this topic, you can ask me on my Discord; I've even created a forum for prompts there and will soon share some. You can also share
yours to help each other with prompting. Now, let's continue with what
else you can do with Gemini. You can also create sheets with Gemini, which
I frequently use, and then export them. It's a useful feature, but if you use it to
perform mathematical tasks, for example, calculating the sum of something, Gemini can
make many mistakes. If you're using it, always provide the numbers upfront; otherwise, Gemini
can still make many errors regarding mathematics. If you don't want to type, you can also record
your messages here. Let me demonstrate it for you. "Hey Gemini" As you can see, Gemini has
converted my voice into text; now you just have to send it, and
then it will provide an answer. A really cool feature is "Image Recognition."
For example, I can send my thumbnail and ask what can be seen on it, and it will describe it.
It's important to note that it's not flawless; there are still numerous errors, but looking
ahead, I envision it becoming incredibly useful, especially for more detailed
analysis of graphs and visuals. I actually wanted to show you Image Creation,
but unfortunately, it's not yet available in my country. However, you can see for yourself how it
works by watching this Google video. Hopefully, it will become available for me too
soon. For those of you lucky enough to be in the USA where you can use it,
please send me some of the images you've generated on Discord. I'd be curious
to see the quality compared to DALL·E. Regarding coding, it's currently limited to
handling minor tasks, offering assistance with small issues. However, it's prone to making
numerous mistakes, so waiting for updates might be necessary. Nonetheless, I recommend giving
it a try to see if it proves helpful for you. I want to leave you with this advice: don't
overcomplicate things and just test and try out everything I've shown and taught you in the
video. I'm confident that with a bit of practice, you'll truly find the best workflow with Gemini.
For everyone who has stayed until the end of the video, I thank you. I've made the free document
available on my Patreon and Discord for you, check it out if you need quick support or
a reminder of how things work. If you're interested in writing uncensored with an AI, take
a look at this video. Thank you for watching.
