Transcripts
1. Introduction: Hi there. Welcome to this
course on Google GMNIF Scratch. My name is Horst, and I'm
instructor for this class. Here in this class, you will
learn different methods and ways to use
Google GM andi for performing various
types of tasks that can help you save time and
automate other tasks. If you're a complete beginner, don't worry because you are going to learn
everything right from scratch without any requirement for
background knowledge. And gradually we will move to
advanced concepts as well. Here in this course, you will
learn to generate images of different types based on the text input
that we provide. So we can generate
images of say, funny cats, we can generate wallpapers or just any
requirement that we have. So we can use these
images for a variety of purposes like creating
power point presentations, using it in a graphic design
portfolio, and much more. Moreover, you will
learn to create text based content such
as articles, email, reports, and other metrics
based on the transcript, such as an interview transcript, group discussion
transcript, and my own. Based on the text transcript, you can drive insights, generate certain things,
we will see how we can take advantage of Gemini
in this situation as well. Moreover, you will also
learn how to generate code based on the image and
also based on the text prop. Here we will see
some examples like creating a login
page consisting of all the STML CSS and JavaScript without any requirement for
background information. If you want to
build your website, mobile application or some
simple coding programs, and you do not have a background in computer programming,
don't worry, you can generate these
things pretty well with the help of Gemini Gemini can help you write
the code as well. And also, if you have the code, you can take an
opinion and review it. So you'll see how we can
leverage Google Gemini for different use cases
and how it could be beneficial for your work. So now, if you're
curious to learn these skills in Google Gemini, you start learning right
now, S in the class.
2. Creating images based on text: Hi friend, welcome
to this lesson. We're going to learn
about how to generate different types of images with
the help of Google Gemini. So let's get started. So right now, you're
familiar with Google Gemini, and let's get into prompting. See here, we are
into the dashboard. We have to just login
through any Gmail account, and then we can
write the prompts. So on the left hand side, you
can see the chat history, and here in the bottom center, we have a panel to
provide an input. So now let's write our first
prompt to generate an image. Create an image of a ginger cat eating
sandwich and hit Enter. Here we are asking
Gemini to create an image of a ginger
cat eating sandwich. It is a creative
image where we have specified the variety of
cat and what it is eating. Let's see the result. Here we have the cat. Here we have a piece of bread
in the mouth of our cat. It is photorealistic image. It's real cat in
the second image, we can see a cat is just near bread and something in between and then we
have the third image, a funny cat relaxing in a park here she has
a plate with bread. Okay. And then we have hungry
cat sitting on a table, a wooden table and with a
plate of bread and sandwiches, and she's describing
her favorite part. Okay, so here is an image of a ginger cat eating a sandwich. We can try generating
more images. Now, create a cyberpunk image. Of London City. Okay, so here we can try generating different
variety of images. It could be photorealistic. It could be in a specific style like we could ask you to
generate an anime style, or the cartoon style. We can also define the mood. Like if you want it to be funny, you can and here, let's see the cyberpunk image of London City. So here we have. So it uses the
color of Cyberpunk, Magenta and different
styles of color. It looks like different
skyline of London and some of the attributes we
have from the London city. Now here we can see
we have some skyline. Although this is not
exactly the real photo, but it is inspirational
or artistically inspired from the
theme. Here it is. Definitely, if you
want to go with any of the images,
you have to use, you can just download it in full size and you can
use it as you feel like. And all this is in
Gemini for free, we have the Gemini Pro, and then we have Gemini Ultra. You can generate the images pretty well in either
of the version, either of the age language
model. So here we have. And if you want to
generate more such images, just hit here, generate more, it will generate more images. So once you are
not satisfied with the images that Gemini
has generated for you, you can always ask you
to generate more images. So in that case, you can decide which
image you want to go. Here, it looks better
than previous. Because here we can see
different skylines. We have the connected bridge and different lighting
conditions. So more variety. In the same way, you
can use these images or we can write some
different prompt like now, create a funny image of a dog. In any way style. Okay, so here we are leaving
Gemini to be more creative. We are not specifying
what exactly we want. We just want to have
a funny image of a dog in the anime style. So Japanese anime are
very much popular, and based on that style,
we want to generate. So here we can see,
these are the images. Let's have a look. It looks
like an oil painting. Okay, so it's good,
not exactly anyway. This one is pretty much
more anime inspired. It looks like. And there's
a beautiful dog here. And in the same way, we have a funny image
of a big chef dog, maybe is cooking something in
the kitchen and here we go. You can definitely download
these images at once. Just go to the Shar tab and here you can click
download all images. Once you want to download
all the images in the prom, you can download all the images. Also if you want to share these prompts with any of
your friends or colleagues, you can just go here, share maybe entire chat all
this prompt and response. You can create a
public link and once this link is created
because for images, it may take some time here
once a link is generated, you can copy this
link and share it on social media or maybe with
your colleagues or friends. It would be very much useful because when you're
generating images, you may want to take an opinion of your colleague or friend. This way, you can share
this link and they can have a look instead of downloading all the images and
then sharing it with them. This way, your
friends or colleagues maybe they can also
help with you the prom. So, for example, create a funny image of
dog in Aimee style. They may ask you to edit it and make a
little bit changes. You can just edit
your prompt and now, create a funny image
instead of dog, you can write a specific
breed German Seafood. In retro anime, die, okay? And here we can. Sing. And now the prompt is changed and we can expect a different
types of response. And based on that, we can
take opinion as well. Okay so now it's retrost
it's more artistic. And here, although it is missing the element
that it is not funny. You want to make it funny. So sometimes Gemini could give you the exact response
that you're wanting. Otherwise, sometimes it may
have this limited capability. And always remember,
Gemini is still evolving, and if you are not getting the exact response that
you are expecting, it will be created in
the coming releases. Here we have successfully
generated different types of images with the help of
Google Gemini and right now, you know how easy it is to create different
types of images. Just experiment with your ideas and generate different images. You're going to learn more
about Google Gemini in the coming lessons till then keep learning and
keep moving ahead.
3. Code Generation- Creating an HTML form based on an image: Hi, friend. Welcome
back to this lesson. Here we're going to
learn about how we can generate code like SDML CSS, and JavaScript from an image. Just like we can provide
input as a text, we can also provide input as an image or a combination
of image Wi text as well. Here we are going to see the code generation
capability of Google Gemini. Now let's start a new chat. Just go top left
corner on your screen, create on the new
chat here you can see we can write a prom just
like we did previously. Also on the right corner, we can see we can upload an image or we can
either use the audio. Now let's upload an
image. Here is an image. Let me show you what
is this image about. This is a hand drawn image
on a piece of paper. Here, there's a written name, a text box, an email, another text box, a checkbox, agreed to terms,
and a login button. So this is a simple image, and we may be wanting to design a website or
a specific web page, log in pH or just anything. And we want to
generate the code. Maybe we do not have a background in
computer programming, and we still want to
generate the code. We can take help
of AI to generate the code and let's see
what is the capability. Let me write a prompt because
if I just upload the image, how would Jamin I know what I
want, what I'm looking for? It will understand the image, but also we need to provide an text prom in addition to it. Based on this image, would you create an STML form with all the SDM, CSS, and JavaScript
in a single web page. And here we can enter. And now let's wait for Google
Gemini to generate a code. Just like we can generate
a code in SGML, CSS, JavaScript, these are the front end web development languages. We can also generate code in different other
languages like Python, Java, Ruby, Icala, any other
programming languages. The back end scripts like parle php asp.net,
different types of code. Or maybe with Microsoft Office, if you want to automate some of your workflow in spreadsheets, Excel spreadsheets,
you can use Excel VBA. Generate a variety of code, and maybe for Linux, you can generate a shell script. And here we can see the response generated is pretty quick, and here we can find
the entire code. Although there's a
security warning, but it also gives you the
Instruction how you can use. It is a basic TML form with labels and input fields
for name and email, a checkbox for agreeing two
terms or submit button, a simple CSS to style the form. And here's a
disclaimer, like note, this is a basic example
and does not include any JavaScript functionality for form validation or
processing the summation, because we have not specified what kind of validation
we are looking. So if we want to
add a validation, we could ask it to do
specific validation, for example, in the email field, we may want at red
symbol or.com. If it is missing, it will not allow you to
enter the email. A name could not be an email
and in the name field, we could not allow a
special characters because nobody in the world has a special characters
in their name. This could be the validation we are looking for
with the JavaScript. Now we can generate it as well. We can improve the code or we can change the
prompt as well. Now let me copy the code first. Here you can just
select it and copy it. Either you can use this
copy code directly. Now let's go to the folder. Now let's create a
new text document, call it, say, sign up. Now let me increase this code. And here the pod is there. Now we need to just rename this file dot STMO
and here we have. Now, let's open. So once we open this file, we can see this is
a web page created. Here we can write name an email. So here we have and this
checkbox is working, and there's a login button. Although this login button, once we press here it
shows a disclaimer, please fill the field. So the name field could
not be left empty. We can write anything like
Andrew or just any pain, you can write and you
can create a log. This way, you can create
code pretty easily. Now you know the capability of Google Gemini in helping you generate different
types of code. You maybe if you're looking
to generate a website, say maybe your website consists
of ten different pages, you can generate a
de pretty easily. Just drawing it on
a piece of paper. And then you can do a
little bit modifications. This web page was
not mobile friendly, we may be wanting to have a
mobile friendly web page. Let's see first whether it
is mobile friendly or not. In order to check whether a web page is mobile friendly or not, if we stretch, it will compress it is
compressing already. So it is mobile friendly, although we want to have maybe specific framework like
BotiTrap or maybe JQuery. If you want to
include these things, just mention these things. If you already have some
background in web development, you can specify these things. You just need to know
how things work. And then you can
generate the code, save your time, and go forward. Also, you can use this
code as a template. If you want to do
any modifications, definitely do the modifications. Say maybe in order to have
the security parameter, and you can also
provide a code for JVu. Here in the prompt, you
can write code and ask JAPT Google Gemini to review your code and
find the mistakes, all the bugs, if there are any. And based on these bugs, you can make the fixes. So here we have seen the code generation
capability of Google Gemini based on
the image and a prompt, we can provide an image as an input and ask
it to write a code. In the same way, you can
also generate a code by just writing a simple prompt write Python code to display some of two numbers
provided by the user. We can write it this
way and based on that, it will generate the code. If you have any functionality, just put it there and
you will get this thing. Also, if you want to
understand the coding or just a specific thing,
you can ask it. It will also explain
you these things. It is very interesting
also in terms of learning and also for
creating the content. You can definitely generate
the code with the help of Google Gemini pretty easily, and this
is how you can do. Take your creativity
to the next level and start leveraging Google Gemini and other AI tools
into the workflow. You're going to learn
more such thing in the coming lessons till then, keep learning and
keep moving ahead.
4. Text generation- email, translation, etc: Hi, welcome back, friends. Here in this lesson,
you're going to learn about how we can generate different types of text based
content such as emails, such as sentiment
analysis and do different types of things based on the prompt
that we can provide. Okay, so now let us
write a detailed prompt. So based on this group discussion,
transcript, write a short summary. Identify or write a
short summary with name of each person
with their roles. And here within codes, I'm going to provide a group
discussion transcript. I can copy it. And here you can see this is a group discussion
transcript, not too long, but
it is medium size. Here we have different
people talking to each other for a
mobile app development, group discussion, and
let's see whether Jem and I could understand
this thing or not. Okay, so here, let me
show you this transcript. So this is a group
discussion transcript. Here we have a project
manager, a developer, UID signer, marketing
specialist, QA analyst, and other person. So they are talking about certain projects and maybe
you are running short of time you do not have such time
to see the video or go with the transcript and you want to generate
a few things, you can provide it as an input. So here you can see, I have
provided this transcript in the prompt itself by just copying and pasting it
and adding a few details. And here, Gemini has smartly identified each
participant with their. And now here's a key summary. So they were focused on
user friendly experience. For app, the development
timeline they have discussed, it is estimated to be 6.5 months including design,
development, testing, and marketing preparation, and additional consideration
they were discussing. And these were the action items, conclusion of the
overall discussion. And the team is aligned on creating a
mobile app development. This is the objective of the
discussion with a focus on user experience,
engagement, and quality. The detailed project
plan will be shared outlining timelines
and action item. So so the background. So if you have a
lot of transcript, or maybe you want to generate
a summary of a book or a long piece of text and you do not have that
much time to read. It would be very useful
to provide gemini and generate the results
and basic things. Okay, now, uh, now
write an email from project manager to Fred informing him about a new join into the team Let's
call her Clark. He have background in building mobile applications and robust coding knowledge. Ask, Fred, what types of tasks can we assign to A. Okay, so now maybe
I want to write an email based on this
group discussion, from a project manager informing him that a new joining
is entering our team, and he has a background in a specific knowledge like
computer programming, mobile development,
and what type of task can be assigned to him? Because Fred is our developer, so that's why we
are asking Fred. So this is the subject
line and the email. So it would very much useful if you want to compose
an email just like that. The subject line
says, welcome to new team member Clark and Delegation
Opportunities. Hi, Fred. I'm writing to introduce
you to Cl who would be joining our mobile app development team as a developer, Clark brings a
strong background in mobile application
development and robust. Okay, so here, based on
the background, and here, as you can clearly see based on the background information
that we have provided, we have not mentioned
the name Natasha. We have just mentioned in
the prom project manager. And because Gemini
is smart enough to understand the previous chat
previous conversations. And that's why it identified that who is the project manager? And now in the email, there's a clear mention. Now, let's try to translate this email to a
different language. Now translate it to French and it will do the
translation pretty quickly. This is email. It will be very much useful if you
have a global team. You have team members working from different
parts of the world. So that's why it
will be very useful. Also translate it to Japanese. Here you can see it just
translated pretty quickly. And here it is, and you can use it as an email and work
in different ways. So this is how we can
work with a transcript. We can provide with
a complex test input and generate different insights. In the coming lessons,
you're going to learn more about Google Gemini and
different other applications. They'll then keep learning
and keep moving there.
5. Report Generation and export document: Hi, welcome back, friends. In the previous lesson,
we learned about how we can generate different types
of text based content, such as the email, finding the insights based on the
given text transcript. Now let's move forward. So here, previously,
we generated an email in English and then translated
it to French and Japanese. And now let's see whether we can export it to docs or not. So in order to export
any document to docs, we can just go here just below the prompt
results of the prompt, just go to share and export. And here you can
export it to docs. Now it's creating a document
where you can work with it and maybe use it as a reporting tool
or some other ways. Okay, so it will open
at dost google.com, and here we have this thing. You can change the font style. You can make it bold, italic or just any other things. And now you can just go
to file and download it. And download in different
formats is also available. You can download in PDF, TX file, plaintext,
or some other format. Okay, so now I have
downloaded this as a PDF and here it is the PDF. Perfect way. Now let us ask you to generate
different types of thing. Create a table. Based on this on the group discussion or wating RACI matrix. Now let's call it create. It is an RCCI matrix. It is a project management
matrix that involves different stages of
work and each person. During different phases here the defining core app features, user interface, developing
mobile application, integrating background
APIs, and these things. These are different stages
of project development, and here you can see
the responsible person, the accountable,
consulted and informed. This is a detailed RACI matrix. Order to create it manually, you will have to spend a lot of time in understanding
these things. So it is better to take help of gemini and generate
different types of wizard. And here you can see on
the bottom right corner, you have the option
to export two sheets. Now let me export it two sheets so that it will
create a spreadsheet, and then I can download it just like we can do with the docs. And here is the sheet. Now if you feel like there's some mistakes of assessment
based on the Gemini, you can definitely work it and modify it,
rectify it manually. Okay, so like the task, responsible, accountable,
consulted and informed, I want to make it
a bit larger like 40 and maybe fill color. You want to add,
apply, and here it is. Okay? And these are the things. Maybe I want to
change the color, the background color to
something else. So maybe blue. Okay. And here you want
to see different stages. And if you want to
group things based on the category like the
design and the development, the development strategies
could be converted to green, maybe a lighter green. The design part, you can convert
it to a different color. Marketing can be converted
to different color. So this will make it more easy to identify
different stages. Social media integration
allows part of the marketing. Okay, and then we
have other things. Okay, so this way, we
can easily identify different category of work
and who is in charge, who is the responsible and
consulted person in each task. Okay, so you can see Gemini made us do this thing in
a matter of few seconds. Okay, so this was very easy
and based on this thing, you can generate
such assessments. Okay, now, Could you visualize the project timeline on
a chart. And let's see. Because the timeline was also
discussed in the project, and maybe you want to visualize it in different
types of charts, bar chart, Sankey or line
chart or other things. And here you can see. Okay, so Gemini has generated a chart. So let's see mobile app
development project timeline. The design UI is
around six weeks. App development requires
12 weeks of effort, marketing promotion
requires eight week and QA testing is five weeks. Although this was discussed, but if we can generate
a visualization chart, it would very much useful in reporting and
for the discussion. Because we all know that
image can speak 1,000 words, generating these reports
could be very much useful. And here is an
image you can see. We can download it
pretty well and use it maybe slide and other things. And if you want to see the code behind generating this image, you can see this
is a Python code. Gemini is running code on the background to
generate this result. So maybe you do not need data scientist for doing such
things or data analysts. Maybe you require for
complex task and evaluation, but if you have to do
some basic assessment, you can do it pretty well
with Gemini in the future, it is going to be more capable. Could you generate
a budget analysis? Report for this project. Okay, and now we are
asking Gemini for a budget analysis report because it would be
also very much useful. Here it is giving
a simple report. You can use it as a
template and you can just change the amount or
add any other details. So it is the disclaimer, it has put on the top. The project is scope
budget categories. The categories could be
the human resource cost of human resource for each
person's salary or hourly wage, the software and tools,
the software licensing. Marketing and user acquisition, like app store listing fees on Google Play
Store or Apple Store. Okay, so it requires the fees, the marketing, campaign costs, social media ads,
influencer marketing, et cetera, and the budget
estimation total budget, next step and conclusion. So it is a detailed analysis. We may want to generate
a report in PDF. So what we can do is first
generate a document. And then once the
document is generated, we can change a
little bit amount, and maybe we want to add some charts or do
some assessment, and then we can print
it in PDF or so. Okay, so maybe we
want to do this thing and let's put a color
in the background. So mobile app development project budget analysis
and reporting, and now let me download it. It's the same way. Just
download it in PDF. And once it is downloaded, you can take advantage
pretty well. Now you can see
based on some input, you can generate different
types of resources, some complex and very much
useful in professional life, and you can do it for a
variety of use cases. This is how easy it is to do. Keep learning and
keep moving ahead. You're going to learn more
in the coming lesson.
Harshit Srivastava, Developer on IBM Cloud, Bluemix
