Transcripts
1. Introductory Video Custom GPT: In November 30 of 2022, this new company that no one has ever heard
anything about it, called Open AI, released its new software to the
public called ChachPT. Two months after the
launch of this program, the active users of the
software surpassed 100 million, and humanity would never
be the same again. This was the first time the
artificial intelligence became so accessible and democratized that anyone
could use it, from students, teachers to employers
and employees. People that understood
the power of ChachPT and were able to
utilize its potential. So exponential growth and their productivity
was unmatched. Almost exactly one year after the launch of PT
three to the public. And the clear
revolutionization of the way that we
interact with each other and produce work. Open AI released perhaps
the biggest update ever in their software. In this new update, Open, I gave us the chance
to be able to customize and release
our own GPT's with absolutely no
coding experience needed the ability
to program and create your own GPT's will give you a huge competitive
advantage over anyone else in your field. Just like the competitive
advantage that people had one year ago when they started utilizing the power of
artificial intelligence. So my name is Lambros and
I have been extremely intrigued by this concept
of artificial intelligence. Open AI and PT, Before Chi three was even
announced to go public, I was one of the first
people that entered the Q list and tested the
beta version of ChachPT. And I have been using
artificial intelligence into every single area of
my life ever since. I've created multiple courses on artificial intelligence
and taught hundreds, if not thousands of
students on how to utilize the power of AI and Ch PT
for their own potential. And I'm extremely proud to
present to you this latest master class on how
to design and create your own GPTs with absolutely no coding
experience whatsoever. In this two hour
course right here, we're going through every single step that
you need to know in order to be able to customize again and create your own GPT. By the end of the course, you will be able to program again your own GPT model
to perform any task that you can imagine completely seamlessly and also potentially listed to the GPT marketplace and generate revenue from it. We're going to start
by analyzing what is open AI and what is 53. Then the differences between
T three and chart four. The updated version, if
you will, of chart three. And then we're diving deeper
into the capabilities of P four and of course its ability to generate custom GPT models. And then in the second
part of the scores, we're diving even deeper
into the creation and customization of those
GPT models on real time. I'm going to be
creating and training my custom GPD model while explaining to you every
single step in the process. And then we're diving
even deeper by uploading this GPT model to the
public and sharing it to potentially millions
and millions of people. So I'd like to welcome
you in this masterclass. I'm going to see you
inside the first lesson.
2. Class Project Descripton: Ladies and gentlemen,
welcome to the course. I'm very excited
to have you here. And in this first
lesson right here, I'm just going to outline the
class project that you're called to complete by
the end of the course. Now again in this
course right here, you're going to learn
amazing stuff about open I by three Ty four, how to create custom GPT, how to list custom GPTs
to the marketplace, and how to gain revenue
from those custom GPT's. I think that the optimal and the best class project
for this course right here is to actually design and configure your own GPT. And paste the link
of your GPT in the class project description
of the course right here, which you can find below. Once you paste your custom GPT in the class project
description, I'm going to be personally
interacting with your custom GPT and giving
tailored feedback in each and every single GPT that you applaud in this class
project description. This is also a great way to
directly communicate with me and engage with
me as your teacher. So I'm very excited
to have you here. This is the class project
that you're call to complete. And I'm going to see you in the first lesson of the scores.
3. WHat is Open AI and Chat GPT-3: Ladies gentlemen, welcome to the first lesson of the scores. I'm very happy
that you enrolled. I'm very happy that you're here. And in this first lesson, we're just going to analyze some basic general principles regarding artificial
intelligence. Open AI, which is
the software company that has developed
chart three, Ch, B four and again
this PT creation of new GPT that we can be analyzing in this
course right here. And we're also going to have
a basic general introduction on what char GPT three is. Right before char P four was
released earlier this year, most of the work regarding
artificial intelligence and again utilizing its
power for our potential, has been done in
chart GPT three. In addition to that,
ChgPT three and Ch GBD four share a lot
of common features. So it is extremely
important for us to analyze some key aspects that are
characteristic of char B three, but are also the same
in chat GB four. Again, thank you very much for enrolling in this course and let's dive into this
first lesson right here. So let gentlemen, we're
going to start with Open AI. What is Open AI and
why did it develop? Chgpt three and Ch B four? Open AI is a renowned artificial intelligence
research organization based in San
Francisco, California. This is the logo of Open AI. It was founded
with the vision of advancing AI technology for
the benefit of all humanity. And this is something
that we're going to analyze a lot in
discourse right here. Something great that
open I has happened, is that it didn't only revolutionize the way
that we interact with AI, but it also democratized
artificial intelligence. Making it completely free
for everyone to use. And also making it able to be used without
any coding experience, which is very, very important. Open AI conducts cutting edge research in
machine learning, deep learning, and
reinforcement learning. Now, we don't need to analyze everything about
machine learning, everything about deep learning, everything about
reinforcement learning. What you need to know is that those artificial intelligence
models created and trained by open AI do not need any coding experience
by the user to be utilized. And this is extremely important because as you can
imagine back in the day, the only people, the only
group of people that could utilize the power of
artificial intelligence were people that studied
computer science. People that knew how to code. But now this is just a fact from the past
because we can all, everybody can
utilize the power of artificial intelligence
without going experience. And what we're going to be analyzing in this
course right here, is going to completely
blow your mind. Now the organization Open, I actively collaborates with other research institutions and seeks to promote artificial
intelligence research as a global public good. Again, they really, really focus on making AI open to everyone. This is pretty much the
name of the company, also OpenAI, they want to democratize
artificial intelligence. They want to make us use
AI in our daily lives, every single one of us. And this is a huge point of the fact that we're able
now to create our own GBTs. Right, We're going to
see later in the scores, it has published numerous
influential research papers and open sourced AI tools to facilitate knowledge sharing
with the AI community. Obviously, we know that
open I doesn't only have GPT and doesn't only give out Ch TPT
for free as a tool. We've got also
generative AI models, such as Dali, for example, that generates image and many, many other AI models right open. Ai has developed a series of increasingly sophisticated
language models, with P three being one of its
most notable achievements. Obviously before GPI has done a lot of research and a lot of experiments with
other AI models. But what made this company
extremely popular back in November of 2022 was
the release of P three. Now this new upgrade, this new upgrade in which
P three became chip four, and all of the things that are inside P four are just
another huge upgrade. And if you have the
information to utilize the power of Chachi
Pt four and creating your own GPT's using
Dally with Ch T P four and all of the other things
that we're going to be analyzing in this course. Trust me, you're going to have a huge competitive advantage over the people
that just use TPT. If you want to make this
connection in your mind, imagine someone not using
artificial intelligence at all and someone using
chat GPT three, right? Someone that uses a three
has a huge advantage. Now, the same
advantage applies to people that use TBT three and people that have ChgBty four and are able to
create their own GBTs. It's just a huge
competitive advantage. So you need to be
ahead of the curve, and this course will get
you ahead of the curve. Now, open AI is dedicated to addressing ethical
concerns related to AI. And is invested in creating safe and responsible AI systems. Now, this doesn't really
apply to our Nie right here. This doesn't really apply to people that want to
be educated with AI. But in general, AI has raised some ethical concerns
regarding all the database. That just subtract all of the
data to give to the users? Right. Some people
have expressed some concerns regarding
where does AI get all of this
information from and who feeds these models
with information. But the more upgraded the
software becomes, the more, let's say, safe and responsible the systems are. Okay, so open? Yes. Research has contributed to advancements in fields
like natural language, processing, robotics,
and autonomous systems. Again, we don't need to focus
at those systems right now. Okay. And it actively
encourages in policy and advocacy efforts to shape AI regulations and
norms on the global stage. This is one of the deepest
desires of this company. Again, to
revolutionalize the way we interact with
artificial intelligence. To incorporate
artificial intelligence to our daily lives. And to make just
everyone being easy. Make it easy for everyone to use artificial intelligence in
every field of humanity, Right, So we discussed
about open AI. What is the desires
of this company? What this company
aims to achieve? Now let's analyze in
the second segment of a first lesson chat, GPT three. Again, chart P three
is the oldest version. The older version, if
you will, of chargPT. I mean older it's one year old. But still the advancements in this field are
very rapid, right? But it shares a lot of
common features with B four. And I think that it's very
important for us to know again some key features of three
because they also apply to B. Four by three. Developed by Open
EI is a state of the art language model that can engage in human
like conversations. It is a GPT of chat. Later in this course,
we're going to be analyzing how to
create other GPT. But TPT is again, GPT focused on language models. It is the third iteration of generative pre trained
transfer, GPT. G stands for generative
pre trained for P transformer series, right? The third iteration and at
three can understand context, generate coherent
responses, and hold conversations on a
wide range of topics. One of the biggest assets
of Chatty three is the fact that it understands
the context, right? And some of the highest
leveraged features of TBT is the fact that when we're talking on the
same chat with the chat bot, it can recall
information that we have previously
mentioned in our chat. So by engaging with Ch TPT
in one chat box, right, In one conversation we can pretty much
something like train, right, the GPT to understand and have more
information about us. So the response that we get, what we subtract out of this
GPD model, where, you know, the closest that
we can to someone that we could actually
get advice from. Right? The same exact thing applies to when we
create our own GPT's. We just give a lot of
information to GPT. We name it, and we have the
best responses possible. So pretty much what
we try to achieve in T three when just chatting a lot in one chat box with TPT giving way more information
than actually needed. So we train GPT in one chat box. The same exact thing
we're going to do when creating our own GPT you're going to see
later in discourse. I want to test about
this for now, right? Ty three was trained on a
diverse dataset containing text from the Internet which makes it highly versatile
and knowledgeable. Now this is why many ethical
concerns have been raised. All right, Due to the fact
that Chap Three has been trained on a dataset from
the Internet, right? And as we know the Internet
is completely free. Everyone can publish
information on the internet. So people started to raise
some questions on, all right, we've democratized
artificial intelligence chip can deliver information
from the internet, but who provides this
information in ChgPTllr? If everyone can upload
something on the Internet and everyone can provide
information to the Internet, then we don't know if this information is
actually correct. And this is something
that open and I mean open IS tries to cover up and
tries again to perfect right, making the information,
the datasets of the artificial
intelligence models being as accurate as possible. Chart three is often used as
the backbone of chat box, chat box, virtual assistance and customer support systems. So again, many websites
have integration of ChgBt. They integrate artificial
intelligence model just again as virtual chat box assistance and customer support systems. And they all run
with chart three. Now most of them run
with Chvty four, which actually requires a high computer
computing power to run. But that being said, both
of them work flawlessly when it comes to just
basic communication and basic service. Again offerings now it has been employed also in
content generation, automatic translation, and
text summarization tasks. Chat PT has been used for
content creation chip three. But as we're going
to see Chatty four, it's just better, right? It generates better images, it generates better content. It has a faster computing power. It also requires
more energy to run. And as we're going to see, it has a cool down
period if you utilize too much of chap four's energy. So you need to
keep that in mind. Don't worry, we're going to. Watch this and we're
going to analyze this in later lessons
of the scores. But for now, know
that both GBT models, both chart three
and chart ty four, have the ability to
generate content. Again, translate
and summarize text. But Ch ty four is just faster
and better than T three. I mean that being said, chart three is
completely free and Chvty four costs something
like ten or 20 bucks a month. But I'm just stating
the facts here, right? Developers, as we said, have integrated chart three into applications for
drafting e mails, creating code, and even
composing music, right? So again, the potential of
Arch three huge, right? And again it is a
language model. It is a generative
language model. So you can imagine what the
potential of ChargPT four is, in which we can
design our own GPT. It also has incorporated
the Dally GBT, which is pretty much
the fact that it can generate images from prompts. So all of these
amazing features we're going to be analyzing
later in the course. Researchers have explored its potential
also in education, healthcare, and assisting
individuals with disabilities. Any field of humanity that
requires information sharing, right, Which is education, which is healthcare, right, And it is also assistance of
people with disabilities, can be hugely, hugely leveraged. Can hugely leverage artificial
intelligence, right? Can hugely benefit from the fact that all of the datasets
of the Internet, all of you know it has
direct access to all of the information that we have as humanity stored on the web, and this is just a huge, huge asset to have charity. Three has also demonstrated an ability to answer questions, provide explanations, and assist with
information retrieval. And this, again, keep
the fact that all of the datasets are stored
online and it has direct access to all of those
online datasets and with a very smooth way and
a very human like way deliver the information
stored on the web to us. So this is just one of the
huge assets of char three. Also, its flexibility and
adaptability have made it a valuable tool for both
developers and end users. Right? This is what
we talked about in the introduction of the
station right here. The fact that people
that have no idea on how to code, for example me. I have absolutely
no idea on coding. I've never taken a
single course on coding, but I have utilized the power of artificial intelligence
and chart three for many, many, if not most, aspects of my life. This past year it has
been absolutely amazing. Huge point regarding
artificial intelligence and chart three is that, as we said, artificial
intelligence is now completely democratized. It is completely
for free and it is accessible from anyone. Right? Regardless of the fact if
you know how to code or not. Right, So open, yes, Decision to offer API access
to chart three has played a pivotal role in democratizing artificial
intelligence, right? Api allows developers and organizations to
harness the power of chart three without requiring
extensive AI experience. What is API? Pretty much
is the fact that you can integrate HPT
in your website. You can integrate this power of artificial intelligence
in your website, usually also
completely for free. This is how open I
has democratized the power of artificial
intelligence, right? And this democratization has
led to the proliferation of AI powered applications and services across
various industries. As you can imagine, various
industries have now become dependent on this power of artificial
intelligence. Due to the fact that, again, it has been democratized
by open AI, which as we can imagine, by the name of a company. They want to revolutionize
the way that we work. They want to revolutionize
the way that people are productive with
AI by their side. Start ups and small
businesses can now leverage AI to compete
with larger players, enhancing innovation
and competition. This is something that you will learn how to
do in this course. Right here, chat
three's ease of use has lowered the barrier to entry for AI adoption and businesses in businesses and
research, right? No coding experience
needed anyone, literally, anyone can use
artificial intelligence in their daily lives, right? It also empowered individuals
and organizations to create AI driven solutions tailored to their unique needs, right? And this has taken a huge, huge, huge upgrade with the
new release of 54. And again, the fact that we
can create our own GPT's, you can see that we will do amazing things when
programming our own PT's. Open AI's, pricing models
and accessibility efforts aim to make AI affordable and accessible to a wide
range of users. It is not that expensive to upgrade from Y three
to charity four. Obviously, it costs some money. But that being said,
the utilization and the usage that you're
going to get out of this, it's going to be extremely more valuable than the money
that it asks per month. That being said, if you
don't want to invest to 54, if you're not ready to
make this transition from by three to Ty four. You can stick around in
discourse to just see if this is something
that you would be interesting into from
these lessons right here. Because I have
bought it and I can demonstrate its value in
this course right here. Now, this democratization has fostered a vibrant ecosystem
of AI applications. Fostering, again,
creativity and information, and open AI's commitment to ethical AI development
remains central in ensuring that the
democratization of artificial intelligence
benefits society as a whole, while mitigating
potential risks. So this is the statement
of Open The company. Now we're done with analyzing
what Open EI is, right? How it democratized
artificial intelligence. What are the goals
of my company? We're done by analyzing the
power of P three and how it revolutionized the way that we interact with those artificial intelligence language models. And I think that
now it is time to move into analyzing
again what TPT four is and more specifically the
differences between Ch B 3.4 because now that we know what their
common grounds is, which is pretty much
everything that TPD does, three does, it is time to move and see what makes by
four so much better. What makes Judge
by four, you know, something that we would pay
for every single month. So this is exactly what we're doing in the next
lesson of the scores, and I'm going to see you there.
4. Chat GPT 3 vs The New ChatGPT-4: Ladies and gentlemen,
now we're done with analyzing P three, the main features of
T three and what made CharPT three so popular and how it pretty much revolutionalized
the way that we produce content and we general are productive
in our everyday lives. In November of 2022, it is time for me
to introduce you to this new Chachi PT
model, ChPty four. All right, by understanding the differences between chart
three and chart y four, you will gain a just
circular experience, if you will, on how to utilize the full
potential of charpy four. Especially in GPT modeling, which is going to be the
next lesson, the scores, So in this again
theoretical part, we're highlighting
the differences between PT three and Pty four. And also we're mentioning
what makes Chargty four so desirable and
so much worth it. So thank you very
much for joining me in the second lesson
of the scores. And let's dive with
the presentation. So charity three
versus chart 54. Let's go. This latest iteration offers improved context
understanding and performance. Right, We're talking
about charity four here, making it a more reliable, first of all, conversation or artificial intelligence
piece. All right. Again, context understanding is something that was very
desired in by three, the fact that artificial
intelligence and by three was able to recall
information in the same chat. This also applies to Ty four but on a more
advanced level. One of the notable
enhancements is that four chart four's
ability to hang, handle lower, longer and
more coherent conversations with the viewers and the users. Again, it can generate
more content, it can generate more text. It is also a chat based
artificial intelligence model. So again, we add props and we chat with chart by
four obviously. But the bigger computing power
that it provides also give us some creative
advantages by four. Can create and generate responses that are
better aligned with the context reducing instances
of non sensual answers. All right? This was
something that used to happen with three, right? But in this case, again,
this higher computing power, which is I think 20
times faster than 3.20, times more reliable than GB, three just generates better
answers compared to charity. Three chit four exhibits a reduced propensity for producing biased or
inappropriate content. If you are a researcher, if you are an educator, right, If you are someone
leveraging the power of artificial intelligence to transmit information to others, you need this information
to be unbiased, right? You need the content
to be appropriate, and this is something
that PT provides. So again, for institutions, educators, people that
conduct research, right? Those people absolutely need the information that
you're getting to be cited and unbiased. And this is something that
T four provides, right? It also demonstrates
an improved aptitude for understanding and following instructions given by users. So again, not only generates
better answers to the fact that the datasets in which it extracts information
from are more upgraded. And again, it's level of
consciousness, if you will, is more upgraded, but it also understands on a higher level the prompts that
we're giving to it. Right? And we've analyzed in many instances that regarding
artificial intelligence, the quality of the prompt, the quality of the
information that you're giving to the program, right, is directly related
to the quality of the answers that
you're getting back. Right. Get four, multilingual
support has been upgraded, making it more proficient
in various languages. This is also extremely, extremely beneficial
as you can imagine. And as we said, the
role of open AI and the key statement of
open AI is that they want to democratize
artificial intelligence. Obviously offering AI and offering all of those
amazing features and all of this computing power just in the language of English
makes absolutely no sense. And isn't that revolutionary. But obviously those chat box have multilingual
support, right? They can generate
again, content, they can generate text, all of the things
that they normally generate in many
different languages. And this also has huge
potential for translators. As you can imagine, B
four's adaptability across different use cases, such as customer support and content generators have
been also enhanced. Right? And this again is completely theoretical what
we're doing right here. But when we launch B
four and I actually demonstrate how much
enhanced those features are, your mind is going to be
absolutely blown, trust me. All right. Content generation, especially with the integration
of Dally, which again Image generated artificial
intelligence model with Bt four has completely changed
the way content is produced. Trust me now, developers
obviously can find tune by four more effectively
for specific applications. Improving its performance in
specialized tasks, right? And when it says developers, it doesn't only refer
to those people in their rooms that know how to code and they're
experts at coding. They also refer to normal people just like me and you, right? And we are able to find tune
ty four to our preferences. We are able to actually tweak
and create our own GPTs, which is absolutely fantastic and in black dozen of those. Course, your mind is
absolutely going to be blown with the capabilities
of this program right here. Ty four also exhibits promise in applications
like code generation, creative writing, and
showcasing its versatility. Not only doesn't it need code
in order to work, right, Not only we can use it without
any coding experience, but we can also leverage its
ability to generate code, to write code for us. So if you are a complete beginner and you
want to code something, but you have absolutely
no idea how to do so, you can actually
chat with Ty four. You can give context into
what you're trying to create A Ty four will
create the code for you. Now, users have
greater control over BD force behavior and output
allowing for customization. That's exactly what
we're going to be analyzing later in
this course with the creation of our own GPT's, which is absolutely amazing. Now, open AI also
continues its commitment to aligning artificial
intelligence systems with human values. Ensuring responsible use. Again, this democratization,
if you will, of artificial intelligence
comes with its threats, comes with its downsides. And the thing is that when we release information
available to the public, when we release those tools
available to the public, which anyone can leverage
and everyone can use into pretty much every
single area of human. You will see that some problems arise just like for example,
the source of information. Again, if the information is
biased or unbiased and open, I really, with this upgrade from such B three to such by four has diminished
a lot of those. Again, liabilities,
data privacy, privacy and security
measures have been bolstered to protect users, interactions with such Ty four. This again is a huge asset
if you are a corporation, if you are Sab
Conductor research. All of those people that
want to have, again, the highest level of privacy and the highest level of security when interacting with those artificial intelligence models, businesses and developers
gain access to a more powerful
tool for creating artificial AI driven
applications. This doesn't really apply
to our case right here. Just mention it because Get four is more tailored
for businesses, is more tailored again
to creatives that want again to create AI
driven applications. And as we can see,
and as we will see, creating our own GPT's
is something like creating our own artificial
intelligence applications. All right, we literally create
GPT robots, if you will. Gpt helpers that help us
and apply the knowledge of PT in our everyday
lives in pretty much any realm of
human endeavor. It is ever able to
be brainstormed. You're going to see in
future lessons now, While 54 leverages
the GPT architecture, it also incorporates substantial model
improvements, right? This deployment
promises to result in more natural and useful
interactions with artificial intelligence
across various domains. So again, natural and useful interactions.
What does this mean? This means that this
artificial intelligence model interacts flawlessly, interacts seamlessly
with the users. And it also has access to an
unbiased database which just retracts information and
delivers it to the users. Open AI's ongoing research and development efforts aimed to further refine and enhance TBT four. This is
something to note. Chpty four is constantly
getting upgrades, right? Constantly
getting updated. And the more people use
it, the more data open I has in order to make it
as perfect as possible. Users can expect more
coherent and context aware responses when conversing with this artificial
intelligence model. Now in addition to that, Bt Four introduces advancements
in language generation, making it a viable tool
also for content creation. And again, content creation
isn't only videos, it isn't only images, it is also text. So it could be blog posts, Twitter posts, Facebook
posts, Instagram posts. It could be images. Thumbnails for Youtube videos. Literal videos can be
executed with Charpy four. This is all content creation. So again, for content creators, Cha GPT and Chifty Four is an absolutely an
absolute game changer. It also has the potential to
revolutionalize industries like journalism and marketing through automated
content production. Exactly what we discussed
about in the previous slide, right? Works and jobs. Absolutely need text in order. You know their literal life
is text journalism, right? Marketing, all of these
are persuasion tactics. Heavily that heavily rely on how we communicate
with other people. And Touchy four is amazing
in communication, right? The model can adapt to various
conversational styles, from casual to professional, catering to diverse user needs. And this, as you can imagine, can be highly leveraged
by marketers, right? That want to switch up the tones in which they deliver information to their
audience, right? So you might have, for example, a marketer trying to sell something in a
professional setting, in a casual setting, in a funny setting, in
a reliable setting. And all of this can be
done with charge Ty four. In addition to that, improved
chat box performance and virtual assistant capabilities
make charge Ty four a more versatile artificial
intelligence companion. Right? As we said, 54 is
designed to facilitate better human artificial
intelligence collaboration in both personal and
business context. Right on top of that, its abilities in summarization
and paraphrasing contribute to more efficient
information retrieval. So again, better datasets to subtract and to
retrieve information, better summarization
capabilities make it a better language model. Okay, the models
language proficiency continues to improve, reducing instances of
nromatical errors in responses. This is especially important if, again, we're atactic research, if we're writing articles
using artificial intelligence. Open AI's commitment to
ethical AI development ensures ongoing efforts to minimize bias and minimize
harmful content. So those are some
huge points regarding by four feedback
mechanisms also and user input are crucial for fine tuning BT four and
addressing limitations. And this doesn't only apply to direct feedback that
we give to open AI, but also some small
recommendations that we can give to the same. Again, chat box to 54. 54 actually takes
into consideration the feedback that we give
and changes its attitude, if you will, towards the user, based on the feedback
that we give. Now this is, ladies
and gentlemen, why you are here. This why you enrolled
in this course. And this is how we're going to navigate through those
lessons off the scores. Again, moving forward from here. This is an amazing slide. 54 offers guidance
and support for users interested in building their
custom GBT like models. This is an absolute
game changer, people that know how to utilize the power of creating
custom GBT like models. We'll get ahead and we'll have a huge competitive advantage over people that don't know
how to do this in the future. And in this course,
we're analyzing every single thing that
you need to know regarding creating your custom GBT models from this point
onwards. All right. Gbd four provides resources and documentation to help users understand the
model architecture and fine tuning progress. And I don't want you to be overwhelmed with
this, all right. B four doesn't need you to know any coding experience in order to create
your own GPT, right? It just also provides
some documentation and just some basic
GPT understanding. All right? In order
for you to do so. But again, with the
information that you're going to be delivered in
this course right here, you are just going to be ready
to create your own GBT's. B four can also assist
users in selecting and preprocessing training data for their specific
domain or case use. We're not going to focus
on this right now. What you need to
know is that again B four encourages
experimentation and innovation. Empowering users to create AI models tailored to
their needs, all right? And those needs are stretched as far as anyone's
imagination, right? We're talking about
every single realm of like human work,
human endeavor. We can create AGBT model
to assist us with it. So this is absolutely amazing. Now, the availability
of GBT four as a resource fosters a more collaborative
and open ecosystem for AI model development, benefiting a wide
range of industries and application and
applications, right? And you're going
to see this again when we create our custom GBT's. How true the statement is and how pretty much every
single industry in this world that
has been created by humanity can
hugely benefit from prompt generation
and prompt again engineering GBT
model engineering. Right? So I think that
we're done with P four. We've analyzed the differences between GPT three and GPT four. And we also, at the
end of this lesson, we also had a sneak peek
on this scalability of GPT four to
create custom GPT, to help users create custom GPT. Now it is time to
answer this question. What is a custom GBT? How can create a custom GPT? And then moving
forward in the scores, we're going to be creating
our own custom GPT. We're going to show you
everything you need to know regarding creating,
again, custom GBT. I'm very excited to move on to the next lesson
of the scores. Thank you very much
for being here and I'm going to see you in the
third lesson of the scores.
5. Introduction to Custom GPT's: So congratulations
ladies and gentlemen for completing this first
theoretical part of this course in which we've analyzed some basic
features of open P three and then the differences between
three and chat GPD four. At the end of the
previous lesson, I've hinted the fact that again, we can create custom
PT's with B four. And this is the pivotal point
in this course right here, in which we move from the general features of open EI again and
P and all of those chat bots into actually
generating our own GPTs. The reason why you
guys are here. All right, I'm very excited
for the next lesson that work that will come and for
this lesson right here. So without any further ado, let's dive into
this presentation, into what are custom PTs and how can we incorporate them in our lives. Thank
you very much. Let's launch this presentation. So what are custom GPT's,
Custom PT's, right? Combine customized instructions,
additional knowledge, and various skills to provide
a unique AI experience. Those are GPT's. Or if you will, language models that we train by ourselves with the
guidance of charity Four. Right. Because again,
we do not need to know how to code to
create those custom PTs. So chart 54, ask us, asks some questions regarding
how we want our GPT to act, what's the tone, what's
the value that's going to be giving, where
it specializes. All of those features we are
going to be demonstrating on a live basis in this call right here in
this course. Right here. To provide a unique
AI experience, users can easily build
their own GPT without, again, any coding knowledge. And I don't know if you
realize how huge this is, The fact that you can
create your own GPT without any coding knowledge, Making AI more accessible
to a broader audience. This is absolutely insane, and a complete game changer. All right, back in the day,
people like you and me, people that don't
know how to code, we absolutely had no shot
on creating our own GPTs. But look at where technology
has gotten us custom. Gpt's can be made
for personal use, company specific
applications, or shared publicly for broader use. This is absolutely insane, because you can create
your own GPT to help you in any realm of
your everyday life. You can create custom
UPTs, I don't know, to even create ring tones for you to wake up
in the morning. You can create a custom GPT that has a fitness plan for you and creates
a fitness plan, a diet plan, a
personal trainer, GPT. Pretty much everything
like music suggestion GPT. You will see that
there are huge, huge, huge applications, right? And obviously, if you're
working on a company, you can also create a GPT that
helps you with your tasks. You can create a calendar, GP. A GPT that helps you
talk to people, right? To leverage your
storytelling skills. All of these can be
done with PT's right. Now the creation process
involves simple steps, like starting a conversation,
providing instructions, and choosing the
capabilities of this model. What I want you to understand is that the process of creating a GPT isn't just the process of us staying in front
of the computer and just coding ending encoding. It is, let's say, a conversation that
we have with such B. Four such four guides
us extremely well into asking us the correct questions regarding how we
want this PT to act. Again, it asks for instructions, it asks for the capabilities
of this program. So it is very easy for every single beginner out there to create and
code his own GPT. Now, GPT's offer functionalities
like web searching, image creation, data analysis, and this enhanced their
ability in various tasks. Web searching is a feature
that TochPty four has, that so it three
absolutely didn't have. All right, You can
paste a web link on Tochty Four and Chit four will
actually access this page. It will analyze
all of the data in this page and provide
relevant feedback, relevant information based
on what the page writes. All right, so web searching is a huge ability of
Arch 54 and it is incorporated into
custom GPT making and also image creation. How is image creation
possible with our custom GPT? The fact that Dali and Dali two, which again are image generated artificial
intelligence models, are incorporated inside
of ChachPt four, which provides the
main computing power of our GPT creation. Right? Custom instructions were introduced to set
preferences in ChachPT, leading to a demand for
greater again control. So again, custom instructions
have been used to code those artificial
intelligence program custom GPT's automate the process of managing prompts
and instructions, making GPT more efficient
and user friendly. So. Back in 53, right, when we wanted to
utilize the power of artificial intelligence and chart 53 to benefit
in our lives. What we did is that we
tailored our prompts, if you will, to be as
spot on as possible. We really focused on
creating the best prompts, because creating the
best prompts, right? Giving the best information to the artificial
intelligence program was the optimal way of actually getting back
the best information. This was the only parameter, if you will, that we
could tweak in order. And when it came to, again, collaborating with the
artificial intelligence model, but now with custom GPT's, we can tweak many,
many other parameters. All right. We can tweak more stuff to get more
relevant answers. All right, to individualize
those GPT models even more. And pretty much this
is exactly what we're going to be
doing in this course. Right here in this course, we're analyzing all of the parameters
that we can tweak and how to understand how to
tweak these parameters to get the optimal
results out of those. Gpt's now the most innovative and effective GPT's
are anticipated to come from community builders expanding the AI capabilities. So two upgrades that
open EI has announced, right to come later this year or earlier in the next
year is first of one. First of all, being able to
create a GPT as a community. So two and more people actually collaborating to create a GPT. And the second huge upgrade is the fact that once you
program a custom GPT, you can actually list it on a GBD marketplace,
powered by AI, and someone that could potentially benefit
with your GPT and doesn't want to
spend the time and energy that you spent
coding this GPT, can actually purchase it, download it, and use it
in his own everyday life. So as you can imagine, being
able to code PT's code, all right, being able
to create those GPT's. Being able to create many GPT's into many different
areas of life, can also open a window
for you to monetize this amazing skill
that we're going to be learning in this
course right here, which is absolutely amazing. Open I encourages diverse
users including again, educators and coaches, to
create PT's democratizing, again, artificial
intelligence tool creation. Ladies and gentlemen, especially
in the coach business, a custom GPT is going to
replace a lot of coaches. All right? Because we can also, regardless of the fact
that we can input websites for custom PTs
to gain information from, you will see that another
huge thing that we can input in custom
PTs are PDF documents. Pdf documents could
be, for example, health data about a person, could be performance
data about a person. Let's say that someone
ran a marathon, right? And we have the complete summary of how that
marathon went. We can actually create
a PDF with the summary, Import it into a custom GPT and create a specialized All right. Gpt, personal trainer GPT. That knows every single
thing regarding this person. Why? Because we've
inputed all of those PDFs with all of this
information around a person. As you can imagine,
a normal coach, a normal fitness coach
could never understand and comprehend all of the data in PDF document regarding
a client of his, but PD can and a custom
fitness code for example, GPT can individualized
information to an to an extreme
extreme level. So this is extremely interesting
and extremely important. All right, the
upcoming GPT store, as we mentioned again
previously in this right here, will feature user
generated GPTs allowing for public sharing and
discovery of innovative uses. So again, if you're thinking about paying the monthly
subscription for GB four, by the way, I'm not affiliated
with Open I by any means. But if you're thinking
about the monthly subscription of GPT four, I would encourage you to
actually go ahead and buy it. Because once you buy the
monthly subscription, what you pay the
monthly subscription. And after discourse in which
you're going to become an expert on GPT
creation, right? You will be able
to create a lot of different PT's for a lot of different realms of
your everyday life. And those PTs can
actually be monetized. So you will be able to
upload those PTs into the GPT store and potentially
make money out of them, which is very, very interesting. Now, GBT creators can
potentially earn again revenue based on the popularity and usage of their custom
GPs in the future. All right, so again, identifying a target audience, creating our target avatar, pinpointing the problems
of this target avatar, and then solving
those problems with custom GPTs is going
to be something very, very interesting in the future. And I'm definitely going
to create more content around this to help more
people understand how to, again, help people
with custom PT's. This GPD store will highlight exceptional GPT's in categories like
productivity, education. Again, promoting
community engagement. So imagine creating
GPT, for example, for studying in medical
school or other universities. Imagine creating a GPT that educates you or helps you
learn a language, right? This is absolutely amazing. And as you can see
in the next lesson, which we're going to be
analyzing the GPT's that. Have been created
by AI, by open AI. And just like demo PTs just
let you know what is capable. In this realm, you
can see that it's going to be very interesting, right to PTs are developed with privacy and safety
as key considerations, ensuring user data
control and security. This is something that we
need to have at the back of our heads that
security again and data protection is one of the outermost important
things regarding open AI. Users can choose to share data with third party
API's when using GPTs. Maintaining, again, control
over their information. Again, you can choose what
information is shared when information is kept
secret with again, OP and I and BD asked
you about the stuff Open EI has also
implemented systems to review GPTs against
usage policies to prevent the sharing
of harmful content. So you can imagine when democratizing something
like the ability to create your own custom GPT can
come with its dangers and people could potentially use
this for a harmful content. This is why OpenI has
implemented, again, those systems to
review the GPTs and protect people from
harmful content. Future developments in
GPT's aim to enable them to perform real world tasks. All right, moving forwards,
more practical applications. This is very
interesting because as we're seeing by
this information, which by the way,
is referenced and is cited by AI's own website. All right, I didn't come
up with these slides. The information on these
slides is directly cited from Open Y's website. You can see that
PD's are gradually transforming from just a
theoretical chat based thing into actually infiltrating
our daily lives. And infiltrating our lives,
making a difference, making real world tasks
and real world impact. As you can imagine, a lot
of people's occupations, a lot of people work, will be completely replaced
by artificial intelligence. So being the person
that can code, the person that can
create those PT's right, you can actually become very useful in the
future society. Jbt's are envisioned as artificial intelligence
agents, if you will, capable of interacting with and responding to real world
scenarios effectively. Those are agents
that we can program, that we can create in every single realm
right of humanity. We can have Dr. agents, we can have personal
fitness agents, we can have personal chefs, recommendations to eat,
recommendations to play. All of this stuff we can create with artificial
intelligence. And the thing is that there's
no ceiling into this. The only boundaries, what
agents we can create are the boundaries of
human imagination. How cool is that? Developers can also enhance GPT's by connecting
them to external API's, allowing integration
with more datasets, more e mails and more. Imagine you a developer
creating a GPT, Connecting this GPT into a
real world dataset that is completely and refreshed
by real world events. And having this GPT, for
example, create news, right? It's like creating a agent, a robot that constantly
2047, right? Creates news based on live
refreshings of pages, which is absolutely amazing. Custom actions can
be defined, right, enabling GPT's to interact
with the real world like shopping assistance
or e mail management. Again, we can actually integrate those GPT's that
we are going to be creating with the real world. We can have a
shopping assistant. We can literally
dictate the GPT to buy something and it will be
integrated, for example, into Amazon or other
online marketplaces to buy and purchase stuff. This is how GPT and artificial
intelligence can have a direct impact in our lives. Enterprise customers can
use GPT's internally tailoring them to
specific business needs, department or dataset. So this is just
another thing that can be done with PT's, right? Tailoring again, PT's to
specific business needs, not only for occupational
usage or for personal usage, but actually in businesses, which is absolutely amazing. Gpt's offer a platform also for businesses to
empower employees and create internal difcilty
intelligence tools without needing coding skills. I want you to imagine a world
in which each business, for example, that has ten
employees, all right? And each employee in
this business has 100 different
difcilitytelligence robots, 100 different artificial
intelligence agents. Coded specifically by him, right, to help him
with his tasks. So it's like a business has 10,000 employees,
not ten employees. If each employee has
1,000 agents, right? Which is absolutely
mind blowing. If you think about this, imagine the competitive advantage of a business that has
trained its employees, right, to leverage the
power of GPT creation. This is going to
be absolutely mind blowing in the future
marketplace, right? And again, businesses that don't evolve towards
this new wave, that don't ride the wave of artificial intelligence and
don't train their employees. To utilize this power of
artificial intelligence will stay behind, right. The launch of GPT's aligns
with open AI's mission to involve more people in shaping AI behavior
and applications. This is the main goal
of Open AI to help people without coding experience get involved with AI creation. And this is also the goal
of this course right here, open AI's approach involves the community in AI development, creating for safer and more
beneficial AI systems. Obviously for everyone, the model picker intensity
has been simplified, right? Integrating various
features like Dali and browsing in one place
back in the day in three. We could utilize
the power of Dali, but you needed to go
on a new website, you needed to input
new prolepts into Dal. It was a completely
new user interface, but now everything is
incorporated in char B four. Right? So we have Dali
and web browsing in one again location
with chB and ch four. This is crazy
because, for example, we could create a GPT model
that suggests images. Or we could create a
GPD model in which we just brainstorm fictional stuff. Creates a complete board
game, if you will, with images that can
be generated with Dally and this is
absolutely amazing. Attachments in TD
also allow users to search through PDFs
and other document types, right, increasing
its functionality. And this is a complete
game changer. The fact that PDF
reading is incorporated in Ft 3.54 chart, it is absolutely amazing. Imagine all of the data
that we can input into those programs and
we're going to have those agents that
we're going to be creating, analyze PDFs into like
split seconds, all right? And gain all of the information. Retrieve all of the information, and give us the best answers to the information
that we have inputed. This is absolutely amazing. Open AI also continues
to update and improve the GPT interface based
on user feedback, ensuring a seamless experience. This is something
that P and I really applies to its work ethics. They always are
seeking for feedback, they always are seeking
for ways to, again, improve their artificial
intelligence language models and this new feature that allows us to create
our custom GPT. Right, so custom UPTs again represent a significant
step in AI, accessibility and
customization, offering personalized artificial
intelligence solution for diverse needs. So I think that
this is pretty much everything about what
custom GPT's are. And now we're done with theoretical part
of the course and it is time to move to the practical part of the
course in the next lesson. Ladies and gentlemen, I am launching my GPT website again. If you don't have B,
if you don't have to. Ty four, this is a
great opportunity for you to check out how the user interface is
to check out the power of this new program to consider if you
would buy it or not. If you have PT and Ty four, this is also an
amazing way to learn stuff and to learn
how to create PT's. In the next Esson,
we're testing some of the pre made GPTs
which were made by open AI and released to
the public to just utilize some ideas and to help us
again brain term some ideas. And give us an example of
GPTs that have been again programmed without any coding to help us in our
everyday lives. So I'm very excited
by the next segment. For the next segment that
is coming off the scores, which is the practical part. Congratulations for completing
the theoretical part. And I'm going to see you in
the next essen of scores.
6. Testing Open AI's Pre-made GPTs: Hello everybody and welcome to the fourth lesson of the scores. I'm very happy that you're here because in this
lesson we're diving directly into the practical
part of the scores. In this lesson,
ladies and gentlemen, we're launching a GP 3.5 and directly
transferring to GP four. I'm going to show you the
interface of chat four. And then we're diving into actually demonstrating the
power of custom GPT's. Because as we said
in previous lessons, Open AI has developed its own custom GPT
to just demonstrate the computing power and
the end list pretty much possibility of this
custom GPT programming. So this is exactly
what we're doing in this as and again in this as we're just demonstrating
and diving through, doing a sneak pick if you
will of the custom PTs that open EI has programmed and
has released to the public. I am very excited to
be the person that shows you this new technology
and this new stuff. So enough with
this introduction, let's move into
actually analyzing custom PTs developed by AI. So ladies and
gentlemen, this right here is the interface of Chat GPT 3.5 All right, and this is the free version of PT that everyone has access to, but if you pay the
monthly subscription, you will have access to P four. And as you can see, GPT
four comes with dally browsing and analysis and limited 40 messages per 3 hours. That's one of the limitations
of chart Pty four. The fact that you can
only if you will occupy a computing power of 40 messages every 3
hours. All right. But it's more than
enough for us right now. We moved and we switched
to charity four. And as you can see, nothing
really happens right here in the chat moving from
3.5 to charity four. But what we're
going to do is that we're going to hit
this Explorer tab. Once we hit this
explota from the menu right from the side barrel that opens and closes like this. As you can see, we move into this area which is the area
of interest in this course, at least right here. This is the area in which
we can create a GPT. And as you can see,
this is so early noticed that it is
still a beta, right? It's still a beta version. All right, We can create a GPT, which we're going to be
creating in the next lesson, but for now, all right, we're going to analyze
those GPTs right here. We're just going
to take a look on some of those GPT's that have been developed by
GPT and open AI. So let's just read the
names of those GPTs. Obviously we've got Daly, right, and it's like the GPT actually directly
communicates with you. So Dally says, let me turn
your imagination into imagery. Data analysis says, drop any files and I can help
analyze and visualize your data chip Classic says the latest version
of Ch four with no additional
capabilities. Right? Game time. I can quickly explain board games or card games
to players of any age. Let the games begin
or the negotiator. I'll help you advocate for yourself and get
better outcomes right. Become a great negotiator. We get creative writing codes. I'm eager to read your work and give you feedback to
improve your skills. We got cosmic dream, right? Visionary painter of
Digital Wanderer, right? And tech support advisor from setting up a printer to
trouble suiting a device. I'm here to help
you step by step. So as you can see, Open AI has developed those GBD models. Just give us, let's say, just a preview of what the capabilities of
this program is. So we got sticker wheez, which is absolutely insane. And I've actually used
it previously, right? So I think we're
going to start with sticker wheez and it says, I'll help you turn
your wildest dreams into die cut stickers, right? Shipped right to your door.
So how amazing is that? Open Y with custom PT's doesn't only limit you in
the website of GPT, right? And open Y, It actually
helps you incorporate other websites or at least navigates you to other websites
to complete your orders. So everything's going to make
sense with sticker wheez. So let's actually
launch sticker wees, So this is the preface
of a custom GPT. So you can see
we've got the image and I'm going to
demonstrate how to actually brainstorm and create your own images for your
custom PT in the next lesson, right, we get the
title of the custom UPT and then the description. I'll help you turn
your wildest dreams into a die cut stickers. Right in dict stickers, shift right to your door. Now this is developed
by PT, but again, when you develop your own GPT, it's going to have your
username right here. So developed by your username. These are some questions
that are preset, if you will, that you
can click to ask. These are some frequently asked questions or frequently asked prompts to help you just communicate with the
GPT in an easier way. And those questions we will
program in the next lesson. So let's see what are the
questions of sticker with. All right, can you make cool
stickers for my laptop? I want a cute manga
style sticker of a cat or create a holographic
sticker of a speed boat. Can you come up with a sticker
ideas for my bake shop? All right, so let's
just come up with the most random
thing ever, Okay? I like cars. All right. Let's say surgery. So let's actually let
sticker, we know that, create a sticker of a let. Por 911, which is a fast car. Right? Let me create this of, let's say of a heart surgeon, right on a 911. Let's say this is the
most random thing that I could come
up with, honestly. Create a sticker of a heart
surgeon on a Port 911. And we've got two
stickers actually, and we can actually choose
which one we prefer the most. Those images are being
created with Dally, which is the artificial intelligence image
generation right program. So let's see, here's
a sticker design of a heart surgeon in a pose 911. It features a surgeon in medical scrubs focused
and determined while driving the
dynamic purse design is prepared for cut
sticker printing right, complete with a solid
white background as a strong black border. All right, and then
I've got this one. Here's another sticker design featuring the heart surgeon
racing in a Porus 911. How cool is that?
How cool is that? I actually love the
sticker right here. Really, really enjoy
the sticker right here. That's so interesting. If you'd like to proceed with ordering stickers
of this design, could you please let me know the quantity and
size you prefer? This will help me provide with the appropriate
link for ordering. Let's see. Let's choose
this one for example. All right, let's say
that looks great. I want like ten of them. Ten of them make them. Let's say, I don't know,
10 centimeters. All right. By 10 centimeters, I'm very
excited by the sticker. I might actually order it.
It looks just very cool. I love it. Great choice to order ten stickers of hard engine raising in a portion
level design, each measuring ten by ten. Please use the following link, and it says order
your stickers here. If we click this link,
we literally go into this page right here in which it is this printing service, right? Sticker printing service. All we got to do is download
this image right here. We can download it from here
D and drop it right here. And we can actually confirm everything else and
buy this sticker. All right, how cool is that? Let's actually go back in the Explore tab and
let's see some other, let's say, examples of open. Let's go to, let's go to
the negotiator, right? I help you advocate for yourself
and get better outcomes. Become a great negotiator, let's say prepare to
negotiate for a car purchase. All right, let's see if we can
become better negotiators. Using artificial
intelligence prepare me to negotiate for
a car purchase. Let's see what the
negotiator comes up with. Great, negotiating for a car
purchase can be exciting, yet challenging to
get started and need more information to
provide tailored advice. Could you please provide
the following details, type of car All right. Are by new or a used car? I am looking for a new. Okay. Race car budget.
What is your budget? Let's say my budget is $100,000 All right. Research. I haven't
done any research. I am just enthusiast.
All right, priorities. What are the priorities? I prioritize, Let's see. Lowering the price and getting
the most value for money. Car. All right. Time line. Is there a specific timeline by which you need to
complete the purchase? I have no timeline in mind. All right, negotiation
experience. How comfortable are
you with negotiating? I am pretty smooth
with negotiating. Let's see. Let's see. So I gave it a lot
of advice, right? Thank you for providing
those details. Here's a strategy to prepare for your negotiation for
a new race car with a budget of $100,000 Prioritizing value for money
and lowering the price. All right, so okay, it gives us the strategy to
prepare for our negotiation. So for example, start by
researching the types of race cars available
within your budget. Look for reviews, forums, and expert opinions
to understand which models offer the best value for money pricing information. Find the average selling price of the models you're interested, or we're not going to
go in too much deal. Because this is not a course
about how to buy a car. But you're getting
the point here. The point is that this
GPT right here is made to help you become
a better negotiator. All of the questions
that it asks, all of the information
that it asks us to give receiving because it wants to become a
better negotiator and this is the strategy. All right. For example,
regarding the initial offer is a start by offering a
lower price than your budget. This gives you
room to negotiate. Then emphasize for
your desire for the best value for
money. Be informed. Knowledge is power
negotiation flexibility. Well, you don't have
a strict time line. Be open to the idea of walking away and coming back later. Right during the
negotiation. Build report. Establish a friendly report with a salesperson.
Focus on total price. Negotiate on the total price of the vehicle rather
than monthly payments. Right. Silence is golden. Don't be afraid of silence
after making an offer, wait for the other
party to respond. Right. Trade ins and extras if discussing trade ins, right? Ensure these discussions don't cloud the negotiation
of the new cars price. And then we got
post negotiation, review the contract, right? Don't feel pressured to
finalize the deal on the spot. All right, and we got
this practice scenario. Let's try a brief
mock negotiation. You're the dealership
discussing a risk car priced at $110,000 How would you initiate the
negotiation to try to bring the price closer to
$100,000 budget? So all right, we're
not going to go into much deal about
the negotiator, but you can see that
GPT gives a lot of information and it has
been programmed this way. So when PT was programmed
in the negotiator, they gave it specific
details. All right. Program the negotiator
telling him hey when, when a user starts
using the negotiator, I want you to always
ask information about the card that he's
trying to buy, right? Then provide feedback on
the research phase on the strategy development before the negotiation, after
the negotiation. All of this is not random. It has been programmed to provide this information, right? This was the negotiator. Let's try one more just for fun. All right, let's
try a more fun one. Let's see, how
about cosmic dream. Cosmic dream visionary
painter of digital wonder. Let's say cat, I've never used. This is going to paint a cat. Yes, it's going to paint
a cat. Okay, great. Let's see what image of a cat
it's going to come up with. Again, this is the logo of, this indicates the fact that
Daly is currently working. Let's see. Cat behold the cosmic fell in a creature of color and whimsy
floating in the universe. Will geometry and
dreams intertwined? Okay, this looks like it paints
illustrations of dreams. Again, let's try the
most random thing, heart surgeon racing on a for 9113 RS. I'm just gigging out
with cars right now, but let's see what
question answers. And also a misspelled heart, let's see if it's actually going to find out that I'm
referring on a heart surgeon. So the first quarter of
generating the image is done. I think that I have a suspicion that it's going to pull
it off. Just pull it off. Let's see what it got.
Okay. Okay. That's dope. And this is actually a 9113
RS which is even more dope. I think he's holding
the defibrillator. He's holding the stethoscope
or the defibrillator? I'm not sure. Yeah,
that's very cool. In a world where
speed meet surgery, our hero in scrubs takes
the wheel of a Port 911 GT, three RS racing through a neon dreamscape that blurs the line between
reality and fantasy. Okay, that's though, like I
might even download this. Yeah, I'm downloading it.
I just downloaded it. I absolutely love this. Okay, so that was cosmic dream. Let's try one more.
Let's try one more. Okay, so let's see which
one we should try. Let's try game time.
Let's try game time. I always had a hard
time explaining games, so I can quickly
explain board games or card games to
players of any age. Let the games begin. Okay, Bet. Explain Monopoly to
a five year old. That's the hardest problem
that I could think of. Monopoly is a financial
game, all right? Let's see if a five year old, how a five year old will
interpret this, all right? Mono is a fun game
where you pretend to be someone who buys and sells places like streets and houses. Imagine you're walking around a big square and every
time you land on in place, you can decide if you
want to buy it or not. If you do, you get to put
a little house on it. Okay? When others
land on your place, they have to give you money because they're
visiting your house. There's also a part
of the game when you go to jail, but don't worry, it's just pretend you can still play and get out of
jail later, all right? And there's a fun
part when you can pick cards and give
you surprises, like winning a race or
having to pay for something. The goal is to collect as
many places and houses as you can and have the most money. At the end of the game,
you win by being the best as buying and
managing your places, so everyone else
runs out of money. Remember, it's all about
having fun buying places and seeing who can make
the most. All right? Pretend money.
That's interesting. That's interesting. That's
interesting. Okay. Let's see. Explain, explain poker
to a two year old. I'm just messing around with a program right now,
just all right. Poker is a card game. Kind of like when you play
with picture cards In poker, grown ups try to get the best combination of
pictures on their cards, like all hearts or old
numbers that are the same. They sit around the
table and take turns, sometimes getting new cards to make their pictures better. They also use chips, which are like colorful to
coins, to play the game. If they think they have
really good cards, they put some more chips in
the middle of the table. At the end, everyone shows their cards and the person
with the best pictures wins all the chips. It's a game of
matching pictures, and guessing who
has the best ones. And remember, it's
just a game of fun. All right. That's pretty cool. Absolutely. I think my favorite one of all
was Cosmic Dream. All right, Which is
absolutely amazing. Let's do a final one
with Cosmic Dream. All right. Create, not
even create a picture. Let's see a course creator. All right, creating a
lesson on chat GPT. This is very surreal.
This is very surreal. Let's see if it can
literally paint my scenario as a cosmic
dream right now. All right, let's see
a course creator creating a lesson on chap. This is the first time that I'm experiencing with this program. Okay, In a whirlwind
of colors and ideas, a course creator delves
into the wonders of a PT surrounded by a vibrant
essence of AI and creativity. If I was an Asian girl, I think this would
be more on point. But whatever, I love
the course one, so I think this is
pretty much it regarding the different types of
GPT's, of custom UPTs. All right, we had a sneak peek on all of the capabilities of those custom DPTs
And now let's move into creating our own GPT. Ladies and Gentlemen,
this is happening in the next lesson
of the scores. In this next lesson
of the scores, we're doing every
single step, right? I'm going to show you step by step how to create
and program GPT, obviously without any
coding experience. So thank you. Verymuch.
I'm going to see you in the next
lesson of the scores.
7. Creating a Target Avatar to aim our GPTs : So welcome gentlemen to the lesson in which we're
going to be starting the process of developing
our own custom GPT. Again, without anything
regarding to coding, right? We need no coding
experience to do the stuff that I'm going
to be implementing and showing in these lessons in this teston right here in
the next lesson to come. Now, before we start with
developing our own GPT, we need to make one thing clear. There are two ways to utilize the creation
of a GPT model. The first is to use this GPT as a personal assistant
and the second way is to actually give
this GPT to others. List this GPT into a
marketplace right, And provided as a service
for others to purchase, for others to utilize. And you can either list it
for free if you want to. In the first case
in which we created GPT to assist our workflow, our work right, we do not need to obviously research
our target audience. Create our target avatar, and visualize the problems
of our target avatar. And how we're going
to solve them simply because we are the
target audience. We are the target avatar.
We know our problems, and we're creating this GBT
to solve our own problems. Now in the second case, in which we are actually
creating a GBT model in order to list it on
the marketplace, that is going to be available. It's probably
available as you're consuming discourse right now, as we mentioned in
previous lessons. Then we need to perform some
basic audience research. We need to create
our target avatar. And brainstorm some
of the problems that this target Adar has, right? In order to be able to create the best possible service or
the best possible product. If we say that this GPT model is a product to help again the
needs of our target avatar. In this right here, we're actually going to
be using Chat GPT 3.5 and not a P four
because as we said, chat Ty four has a cap of
40 messages per 3 hours. So we have absolutely no reason to captivate all of the energy
and utilize the energy of Charbty Four when
we're brainstorming a simple target audience
research process and a target avatar creation process which is extremely and basic. And I'm going to
be showing how to do this in the
Sasson right here. So we're using Chat GPT 3.5 to brainstorm
our target audience, to create our target avatar, to identify the core problems
of our target avatar, which again we are
going to be solving with the creation
of custom GPT's. I hope that this
introduction made sense. Again, if you want to create
a custom GPT for yourself, then this lesson right here isn't necessary
for this process. You don't need the information
that we're going to be delivering in this
information right here. That being said, at some point, you will I guess, create a GPT and list
it on the marketplace. Because it is a knowledge that you have and you have
bought with the scores. So it's good for you to know. So every time to
create a product, we need to identify
our target avatar. Identify our target audience. Create our target
avatar and then obviously brainstorm
its problems. Which we're going to be doing
in this lesson right here. So let's launch by three and let me show you
exactly how to do this. So here we are, ladies and
gentlemen inside chat GPT 3.5 So the first thing is let's take a hypothetical
scenario, for example, that we are creating a GPD model for something that many
people can utilize, that many people
can be served by, let's say that we're
creating a personal trainer, GPT. Okay, let's see. I am creating a virtual
personal trainer. I want to identify the target audience that
could benefit from it. Let's see what activity replies. Creating a virtual personal
trainer is a great idea, and identifying your
target audience is a crucial step in the development and
marketing of your product. To determine your
target audience, consider the following
steps, okay? Market research,
demographics, fitness goals, fitness levels,
lifestyle and interests. Pain points and challenges, all of that stuff. Who do you think would be right? Benefit more from a virtual personal
trainer, right? Who should I target
with the service? Let's see. So the
effectiveness of a virtual personal trainer and the ideal
target audience for your service can vary based on the specific features and
capabilities of your product. However, here are some
groups of people who may particularly benefit
from a virtual trainer. So we got busy
professionals, right? Beginners, homework,
and enthusiasts. Individuals with specific
fitness goals, right? Seniors, travelers,
and individuals. Health limitations. All
right, that's interesting. Budget consumers, goal oriented users, socially
conscious consumers. Let's try to pick a more
narrow category of people. We can go either with busy
professionals which I like, and homework at enthusiasts
which I also like. Let's go ahead and say, okay, I will narrow my virtual trainer to work out enthusiasts. Let me change this.
Enthusiasts create a list of people that could benefit from it. I narrow down to
homework enthusiasts, who are homeworker enthusiasts, Stay at home, parents, remote workers, homeowners,
apartment dwellers, elderly individuals,
fitness beginners, individuals with
time constraints. All of those people right here, people with social anxiety, weight lifting, hobbyists, technology enthusiasts,
all of these guys. I'm not going to get into much
deal again because this is a very specific example
that I'm providing right here of my target audience. But now let's actually
create our target out. I want you to create my target avatar. Let's say I want you to create the target avatar of this group of people. Make sure to describe
his day and why he belongs in this group
again of people. All right, so certainly let's create a target out for the group of home
worker enthusiasts. Sarah is 32 year old
professional who embodies characteristics and
lifestyle of this audience. We got Sarah. She's 32. She's a marketing manager, married with two young children. Right, living in a
suburban neighborhood. So we got the description of
Sarah, her morning routine, her work, her evening routine, her fitness goals,
right, and why. Sarah belongs in the Home
workout enthusiast group. So she belongs to this group because her daily
routine is filled with responsibilities and she values the convenience and flexibility
of exercising at home. So we could create a GPT on Sarah and we
will create GPT on Sarah. But now we have undergone all of this process of mentioning PT, the fact that we want to
create a virtual assistant, then moving on and researching
the target audience, narrowing down to people
that are home enthusiasts, homework and enthusiasts,
and then creating. Sarah, we did all of this process to lead up to
this prompt trade here. And this prompt is mentioned
the ten biggest pain points, right, Of, of my target avatar. All right, and based on
those ten pain points, those are the ten pain points. All right, to rephrase it, that we will try to solve
with our custom GPT. Right here are the ten
biggest pain points. We got time constraints, we got the lack of motivation, we got the limited equipment, we got inconsistent routine, we got family interruptions,
fitness plate, lack of accountability,
safety concerns, variety and boredom,
nutrition challenges. Those ladies and gentlemen, are the ten pain points, the ten biggest pain
points that led to Sarah, our target Avadar, to become
a exercise enthusiast. We're trying to enhance this
process of exercising at home with the Virtual Assistant GP that we're
going to be creating. And those are the ten pain
points that we want to solve. All right, now I want to, for example, we could say, elaborate on pain
point number one and suggest five solutions that virtual assistant could provide, a virtual, let's say, training assistant
could provide. Let's see, certainly. Let's elaborate on
pain point number one, which is time constraints, and explore five solutions that a virtual training
assistant could provide to address this
challenge for Sarah. So first of all, we got
time constraints, right? This is the pain point
elaboration Star is busy scheduled as a
working professional, a mother of two children,
leaves her with her limited time for
dedicated workouts, solutions, sort and
effective workouts, customized workout plans, schedule reminders,
workouts, anytime, anywhere, time management tips. So this is exactly
how we're going to program our DPT based on those solutions for
the first problem out of the ten core problems
that our target avatar has. So we want our custom PT to provide short and
effective workouts. We want our custom PT to provide customized
workout plans. We want custom PT to provide scheduled
reminders for workouts, suggestions for workouts
anytime, anywhere, and on top of that,
time management tips. So this is pretty much, ladies and gentlemen,
how we provide. First of all, we brainstorm the category of GPT
that want to create. All right, we brainstorm
again our target audience. We narrow down to
our target avatar. Using such PT, we come up with the ten biggest pain points
of our target avatar. And then based on
those ten pain points, we get each and every
single one of them and we provide solutions to
those pain points. All right, so this is absolutely necessary
if we're creating a virtual product and a custom GPT is a virtual product that we're
going to be selling online. So being able, all right, to program your custom PT
and identify the needs of your target avatar
before you list your GPT online for sale is, as you can imagine, of
utmost important importance for this job right here. Again, if you're
creating your own GPT, you don't need to go
through the stuff. But I think that the information
in this lesson was easy to digest and it
was fairly fast. Now that we're done with researching again,
our target audience, narrowing down to
our target avatar, identifying the
ten core problems, and suggesting solutions
to those problems. It is time to launch such BD
four and create our own GPT. So this is exactly
what's happening in the next lesson
of the course.
8. Creating our Custom GPTs: Harry Barn. Welcome to the
sixth lesson of the scores. This lesson right here is the exact lesson that you
have been waiting for. In this lesson, we are literally going to
be going through the whole process of
designing our custom GPT's. All right, we've described in previous lessons everything regarding artificial
intelligence, open AI three, P
three versus P four. Then how to brainstorm
our target audience and create our target Avator based on what our custom
GPT is about to deliver, what service it is
about to deliver. And now time, ladies and gentlemen to dive
into this process. I'm very happy that you're here. Let's launch B four and let
me show you exactly how to customize your own
custom chat GPT model. This again, ladies
and gentlemen, is the interface of open AI. And we've got GPT 3.5 right now. Let's switch to chat GPT four. Now we're going right
here in the Explorer tab. We're going to
close the sidebar, and we have all of
those GPTs that we explored in previous lessons. But now we're not interested
in this part right here, we're interested in
this button right here, the Create a GPT button. Let's go ahead and
create this GPT. Now, as you can see, the interface that
we've got right here is divided into
two subsections, the left part and
the right part. In the left part, we've got the so called GPT
builder and two buttons, the create button and
the configure button. Right, We will be going through
every single button and every single aspect of the left and the right
tab in this tess. And I don't want you
to stress about this in the right tab
as we're going to be giving information in the GPT builder and as we're going to be
building our GPT, again without code code, just by answering the
questions of the GPT builder. In the right tab,
we're going to see our custom GPT model
being created. All right, and we can
actually engage with this custom GPT model that we
have created in the right. Again, part of the screen in the left as you can see here, again, we can
message GPT builder. In the right we message our GPT. Now look at this, this ban right here is a button in which
we can also attach files. All right, and this is
one of the biggest, again aspects and biggest
selling points of this new GP builder
update the fact that you can attach PF
files. All right. You can attach guides, more information
regarding what your GP, what you want your
GPT to be doing. Right in the GBD
builder, which is, you know, absolutely
mind blowing. All right. Now let's go ahead and actually engage with the GPT builder to show you all the capabilities
of this program. Let's say again that we want
to create a GPT which is going to be a for enthusiasts. This is our GPT, a personal trainer for
home fitness enthusiasts. The GPT builder says, hi, I hope you build a new GPT. You can say something
like make a creative who helps
generate visuals for new products or make a software engineer who
helps format my code. What do you like me
to make a personal trainer who specializes
in workouts? Make a personal trainer
who specializes the specialized
in home workouts. You can see that it
is updating the GPT. And once you see this, again purple color, this means that open Ye utilizes
power of other software. For example, you remember
that the same purple color we would see if we were
utilizing the power of Di to generate images. Now we also creating new GPT behaviors
and this is why we have this purple image. First thing is that you
can see that the GPT, the preview part of the GPT, has been updated to your personal trainer for
effective home workouts. And we also have our first
questions that you could ask, the predetermined
questions that we could also tweak in the configured
tab later in this video. So just a full body
workout for beginners. These are just some
questions that are GPT model that
we have designed. All right, Wants users to be asking in order to engage in a better way
with the GPT model. How can I stay motivated
to work out at home? Give me a 20 minute cardio routine nutrition
tips for muscle gain. All right, so let's move actually to the
GPT builder again. How about naming this GPT
home fitness coaches. Does this name work with you? Now we're brainstorming
the name of our GPT, and I honestly think that
home fitness coach is an amazing name,
let's say. Sure. Let's keep home fitness coach, okay, let's keep home
fitness coach amazing. Again, we're generating
the profile picture now. All right, because we
want the profile picture, we have updated the name. Now it's time to generate
a profile picture. And how we're going to do
this, we're going to do this obviously with Daly. All right. This is okay. So Daly is the artificial intelligence program
which generates images. And this is the image
that was generated. So here's the profile picture
for home fitness coach. Do you like it or do you like? To make any changes. I don't like it that much. Okay. Let's make it more minimal and focus
on muscle, right? Obviously we can tweak the profile picture as many
times as we want, right? It doesn't matter. But again, the profile picture isn't
that important in GPD. I just didn't like this
because we're talking about a home fitness coach, and this is not home, this is a three in a heel.
I didn't like it that much. Let's say it's generating
another profile picture, now that's better. Here's the updated
profile picture. More minimal and
focused on muscle. What do you think? Would you like any further adjustments? No, I love it. So we're keeping
this and this is the interface of our GBT, right? Our profile picture
will be created. And your personal trainer
for effective workout. Now we could be able to, again, talk with our GBD model. Right from now, right,
we can message our GBT, but I want to wait
and actually tweak it more with the GBD builder.
So it's as great. Let's refine the context
for home fitness coach when thinking about the advice
and routines it provides. All right, are there
any specific types of workouts or fitness goals
you wanted to focus on, like weight loss, muscle
building, or flexibility? It's actually a great question. Are there any specific types of workouts or fitness goals
you wanted to focus on? No, I don't want it to focus on specific workouts. I want it to provide tailored personal training in every to every
single user, right? To every single user. Let's see. Let's see what the GPT
builder is going to answer. Again, obviously we
don't want to just have one preset type of workout. That being said, we could
create many different GPT'sor. Could create a GPT for example, for weight loss or GP
for muscle building, or GPT for flexibility. But this is just a general
that covers everything, right? The general GPT that covers
everything. So let's see. It updated the behavior
of the GPT and now it is mentioning the fact that the behavior is updated
and what will we gain. So great, home
fitness codes will provide tailored personal
training to every user, adapting to their
unique fitness goals and available resources. Next again, let's think about the tone and
styling of interaction. How do you envision home
fitness codes communicating? Should it be strictly
professional or more friendly and casual? Any specific language or phrases it should
never use or avoid? All right, we're moving to the second part
of GBT building, which is actually, let's say, the personality of our GPT. How do we want our GPT to be
communicating with people? All right, again, this
back in the day is used to require code
in order to be done, but now it just can be done with a simple conversation
with the GPT builder. I want the fitness. I want the GPT to be very energized and motivational when interacting with the users. I want it to make them want to work out even when
they don't feel like it. Obviously, we're referring
about a home workout routine, so we want a very
energized GPT to keep those users energized. Let's see what the GPT builder
will come up to right now. Obviously, we're going
to be interacting with the GPT later in this lesson to see how good we actually
program the GPT. Let's see, it is updating
a GPT right now. And once it is done, you can
see GPT behavior updated. So perfect home fitness
coach will interact in a very energized and
motivational way aiming to inspire
users to work out, even on days they may
not feel like it. Now let's make sure everything
is as you envision. Try out home fitness coach in the playground to the right. Right. If you have
any refinements, let me know and I'll
adjust it accordingly. How does it sound? So
it sounds amazing. And let's go ahead and test our GPT that we created right now in less
than 10 minutes. So again, this is our GPT energizing personal
trainer for home workouts. And now that we
add the fact that we want it to be energized and Motivational, you
can see that it changes the whole
layout of the GPT. So it changed the
subtitle and it also changed the suggested
words that we can use. So for example, give me a
workout to boot my energy, quick morning workout routine. How to stay motivated
for daily exercise, healthy eating, eating
tips for active days. All right, so let's
start with one of those, again, preset questions. So give me a workout
to boost my energy. Let's see what our
custom GPT will answer. So absolutely, let's get that energy up before I
tailor a workout for you. Could you get me, could
you let me know a bit about your current fitness level and any specific
goals you might have? This will help me create a more, the most effective and
enjoyable workout for you. Now, this, ladies and gentlemen, may seem like a random
answer given by our GPT, but if you note what we
mentioned in the GPT builder, you will see that this
is not random at all. All right? Because
I asked, all right. I asked the GPT builder
to program this GPT by providing tailored
personal training to every single user. And this is exactly
why this GPT is asking is asking me information about my
current fitness level and specific goals
that I might have. All right, It wants to give me tailored advice and this is exactly why it
is asking for it. So this answer is completely based
on the programming that we did in the
GPT builder, right? And it's also energized
because it says absolutely, let's get that energy up. Okay, so let's actually go ahead and answer to this question. Before I tell you
record, could you let me know a bit about your
current fitness level? So, I am a male, let's see, 70 kilograms, okay? I normally work out
three days per week. I work sitd on a chair. I have let's say
20% fat in my body. Let's see. Great, thanks for sharing your details based
on your information. It sounds like you're moderately active and looking for a workout to
boost your energy, especially considering
your sedentary job. I'll design the balanced routine that combines
cardio strength and flexibility exercises to enhance your energy levels sustainable
for your fitness level. This workout can be
done at home without any specialized equipment.
How amazing is that? We literally right,
we literally replaced a tailored personal trainer with our custom GPD model in
less than 15 minutes, which is absolutely insane. I'm getting very
excited by this. So this is our
energizing workout plan. We've got our work warm up, which is 5 minutes
right march in place. Arm cycles, leg swings,
dynamic stretching. Then we've got our cardio
session, which is 10 minutes. We've got jumping
jacks in knees, butt kicks, shadow boxing rest
and repeat the cycle once. Then we've got our
strength training. Push ups, squats, try
dips, plank rest, flexibility and
cool down forward, bend catcutretches,
shoulder stretches, deep breathing, and again,
some tips for success. So stay hydrated.
Listen to your body. Consistency is key,
and nutrition, right? Ensure that you're
reading a balanced diet to support your energy levels. So let's actually, and also we've got our message
right here which says, remember it's more to start each workout session with a warm up and end with a cool down
to prevent injuries, right? Let's get energized
and conquer the day. This again, R GPT, which we told it
to stay energized. So let's actually create a weekly routine. So
it says absolutely. Building a weekly
training schedule that targets all muscle group is a great way to ensure
balanced muscle development. Avoid overtraining
since you're focused on muscle building would include
a mix of strength training, cardio, overall health, right, rest days for recovery. So let's see, Monday we're doing upper body strength Thursday, lower body strength Wednesday, active recovery, right,
activity, light walking, yoga, stretching, goal
enhanced flexibility, and eight muscle recovery. Thursday, back and
core strength. Friday, full body workout, Saturday, Cardi and core, right? Sunday rest day, and
some additional tips. This is absolutely amazing. And it also, again, has us being energized at the
end of the conversation, which is absolutely amazing and kind of wholesome right now, we could add more stuff
to the GPD builder, right, To tailor this
even more to our case. And this can be done with the configure Tap in tab in the configure tap
which changed the name, for example, which could
be home fitness coach. This is the name and
it should update here. Right, And we've got
the description. Let's could we reload this? No, not exactly. Okay.
In the description energizing personal
trainer for home workouts, these are the instructions
that we're given to our GPT. Those instructions
have been created. All right. By the
GPT builder based on the information that
we provided in our chat. So based on the information
that we provided in our chat, the GPT builder
says, for example, the role and goal of
the fitness coach is home fitness coach
is a dynamic and motivational personal
trainer for home workouts. It offers personalized
fitness advice catering to various goals and fitness levels and is adept at crafting diverse workout plans and nutrition
advice constraints. Avoid suggesting workouts requiring specialized
home equipment. Right? Those are some
constraints that GPD builder has
brainstormed by itself due to the fact that we mentioned
that this is a home workout GPT prioritize safety and adaptability in workout
recommendations. So those are also, we've
got more, All right, We've got guidelines
capable of diversifying, diversing workout plans from card to strength training for
different fitness levels. Then we've got some guidelines
and also personalization. Communicating in an energized
and motivational tone, inspiring others to exercise even when they lack
motivation, right? Use uplifting and encouraging
language to boost enthusiasm at commitment
to fitness routines. Those are the conversation
starters, right? Those are the prompts that we urge users when
they're communicating and interacting with
UrGPT to use to have like a great conversation
start just to start again, communicating with RGBT'vet as conversation starters give me a workout to boost my energy. How to start motivated for daily exercise or quick
morning workout, routine, healthy eating tips for
acting days or let's say create a Pero analyzed. All right, nutrition
schedule for me. This could be another
conversation starter. All right? Then moving on, we've got the knowledge and
in the next lesson, which we're going to be be
configuring even more our GPT. We're going to inspect
this knowledge tab. Right here, you can see we've
got an upload file stab. And this upload files
Stab gives us, all right, the option to upload
PDF documents and more. Pretty much videos,
photographs, PDF documents, Everything we need our
home fitness coaches to have incorporated
in its knowledge, in its computing power to provide the best
assistance to our users. In the next lesson,
we're going to be downloading PDFs
from the web, right? Specialized PDFs for fitness
coaches from the web. And uploading here and
testing how dar GPT reacts. Right, Then we're moving
to its capabilities. This is, if you will,
the purple, All right, the purple line which is loading
in our GPT creator mode. Capabilities are web
browsing, dally, image generation and code
interpreter interpreter. Right? Web browsing gives us
the ability to paste a link, again, a page from the web
in our GPT, and this link, this gives the
ability of our GPT to browse the web and
gain information from the web without
having to input, All right, our information
inside the GPT. Right then we've got
da image generation. We discussed about what
damage generation is. Is it pretty much the ability
to generate images from prompts that we give without
exiting the depplication. And then we've got
code interpreter in which pretty much it can analyze data that we give it and showcase the data with
different graphs and stuff, which will be very effective, especially in a
fitness planning code. We could give data, for example, from our runs, all right? And it could plot
the data and give us more information
regarding our runs. Obviously, we've
got more actions. And the more actions the ad actions option is
if we want to incorporate even something more
specialized that we can do it with our own code. Obviously, it is
possible to create your own GPT model without
knowing how to code, right? But if you know how to code, you can actually
incorporate your code in your own GPD model to
make it even stronger. All right, so ladies
and gentlemen, this was our first interaction
with the GBT builder and our own custom personal fitness codes
for home workouts, GPT, that we have created. Let's move now to the
next lesson in which we're going to be
diving even deeper and configuring our GPT
even more by inputting information and
PDF documents that we have downloaded from the web. Again, to expand the
knowledge of our GBT. Thank you very much for
being here. I'm going to see you in the sixth
lesson of the scores.
9. Configuring our GPTs: Ladies and gentlemen,
welcome to the 7,000 of course in
which we're going to be testing and refining our custom GPT custom
fitness coach. In this lesson re, I'm
going to show you how to import PDFs and
how to download firstly PDFs from the web to import them inside
your custom GP in order to train automatically your GPT into any information
you are going to give. And also I'm going to
show you how to refine your GPT and add some
more specific details, some more specific
information in the configured tab of, again, our custom GPT to help just
with the user experience. All right, so welcome again
to lesson, let's dive. Let's open chat GPT. Let's open GBD four. And let's actually tweak even more parameters
out of our custom GPT. So this ladies and
gentlemen is right here. Again, the window that we have. At the left side, we've got
our GPT builder, as we said, and in the right side
we've got our custom GPT. Now in the previous
lesson we have configured some very basic stuff
around our GPT stuff like, for example, the name,
the description. We've got our
profile picture and the instructions of our GPT which are not going to change. That being said,
let's now analyze how can we dive even deeper into the configuration aspects of
our custom GPT and what are the next abilities that we
can add into this custom GPT? So the first thing that I would like to tweak with
you is the knowledge, All right, additional files
for this GPT to reference. Sometimes, especially if we're creating GPT's around fitness, we want our knowledge
or the knowledge of the GPT that we're creating
to be evidence based. All right, especially if
we're creating, again, GPT's that have to do with
health of individuals, we want our staff to
be evidence based. Now, I'm not here to question
the database in which open AI and GPT gets the data
to deliver to the users. But sometimes we need
to be extra careful. That's why in the case of
our home fitness code, I want my custom
GPT to be getting its knowledge from
a specific source that I have dictated. This specific source is, for example, this
book right here. This book is the Primary Personal Trainer
Certification course. Okay, The third edition. And it is 186 pages,
as you can see, of every single thing that a personal
trainer should know. Anything from anatomy
to physiology to how to treat people
to everything, right? Obviously, training programs, the usage of
proteins, all right? All of the stuff. We also have test quizzes. Look at this. All of this is just information that a personal
trainer should know, 186 pages and I want my program, the whole fitness codes to be trained based on
those 186 pages. So what I'm going to
do is I'm going to go here and download this. All right. Where downloads save. Now that our PDF is saved, it is time to import it as knowledge to
our customer, PTs. We go to here, upload
files. All right. Desktop or download
Personal Trainer again, bookmark, load this. And as you can see here, you can see that the
PDF is being uploaded. So we're going to
leave this here to be scanning again the PDF
document as it is uploading it and it is implementing all of the information
that pretty much is stated inside the PDF it is utilizing the PDF is uploaded. All right, so check
this out now. Let me go to
configure. All right. After this I'm going to add, every time you suggest a, let's say a program, I want you to reference it. Let's actually add
this to the chat box. The chat box and not in the configure menu.
It's good to create. Let's say every time
you suggest a program, I want you to reference it from the PDF I have uploaded. Let's see what the GPT
builder is going to say. It is, first of all, updating our GPT to do exactly
what we state. So to be 100% sure and positive that the
home fitness codes doesn't just come
up with random, let's say, fitness
advice that aren't evidence based and
aren't from our PDF. This is why I added this prompt. So look at this, let's see, let's see what the GPT
builder comes up with. I mean, it is pretty
specific what we're asking here and
it's still loading, so I'm very eager to see if it's going
to respond to this. Let's see if everything is
fine in the configured tab because I'm not
seeing it respond, so it looks like it's responded. It's doing one more time. Again, it is loading
one more time. This prompt, I mean I get it
because it needs to analyze the whole article and then to be able to suggest
every single time, the same to reference every single time when
we have a workout. Here is the GPD builder. After again, I've set
this prompt one more time because it had a rough time at the first time to
interpret what I'm saying. But now it says your request
has been incorporated. Now, every time I suggest a workout program
or nutrition plan, I will reference it directly
from the PF you've uploaded. If you have any specific
fitness goals or questions, feel free to ask and
I'll tailor the advice based on the information in
your PDF. Let's get started. Let's say, for example, suggest a five minute energy boost
workout for upper body. Let's see what the home
fitness coach comes up with. Great choice for a
quick energy boost. Here's a five minute upper body workout that you
can do right at home. This workout is
designed to engage your arms, shoulders, chest, and back, giving you
a nice energy kick and strengthening
your upper body. Remember, proper form is key to avoid injury and
maximize the benefits. We got arm cycles. All right, we've got push ups, begin in high plank position with your hands slightly wider. Then we got tricep dips again. This is 1 minute, 2
minutes, 3 minutes, 4 minutes, right, and 5 minutes. Let's see if it is going
to reference our PDF. If this workout,
this quick workout, doesn't boost your
energy levels, it also helps improving muscle tone and durance in
your upper body. Keep up and you'll
notice improvements in your strength and
energy over time. Remember, it's all about
consistency. Let's keep moving. All right? It didn't
actually reference the PDF. All right. Let's actually go
here in the configure tap. All right? I wanted
to say every time. All right, you reference
for the PDF I uploaded. Let's actually add it
in the configure tab. Let's actually add
the same exact prompt in the configure tab. All right, let's see
if it's actually more reliable to add prompts in the configure tab rather
than the create tab. All right, that's going
to be interesting. Let's first of all wait
for it to update the GPT. I think that's the first
thing that we've got to do. Then I have some new
plans. All right. Another suggestion that I
want to tweak in my GPT, which is to actually add images, all right, every time it suggests a different
type of workout. So let's see updating GPT again. It takes time and it takes
computing power because four, as we said, it requires
a pier computing power, the g 3.5 And maybe
we're asking a bit too much from the program right now, but let's
see how it goes. Okay, let's see. Quick morning workout routine,
Searching my Knowledge. That's very good.
This means that it's probably searching the
PDF file right now. And I think that we
solve this problem if this is what I think it is. But I think this right here, this icon that you
see is the code, I'm sorry, the code interpreter. And it interprets the
PDF file right here. Let's actually see what
it's going to come up with. I think that the
solution that we needed to give to this
problem right here is that we needed
our PDF prompt, which is every time
you suggest a program, which is the reference from
the PDF I have uploaded, we needed to be in the
configured tab right here. All right. Not just
in the Create tab. Let's see what the home fitness coach is
going to come up with. Again, this requires
some computing power, and as you can see, it
is taking some time. But that being said, we are
programming a whole GPT here, which, you know, a month
ago or two months ago, it would be pretty much
impossible to do so without any coding experience, there's no problem for us
waiting to have this done. So what happened here is
actually very interesting. All right. It says that
I've searched through my knowledge source for a specific quick morning
workout routine, but I wasn't able to find a
directly relevant section. However, I can still create a personalized morning
workout for you based on general
fitness principles. So how absolutely
amazing is this? And this is why it took so much time for it to
answer it is because it literally went through
186 different slides. And it realized that, first of all, there is
no morning workout here. And second of all, all right, we need equipment to
do these workouts, which is something that
we've mentioned right here, that we need to avoid suggesting workouts requiring
specialized equipment. So as you can see, two of
the problems that I've given to my GPT go against each other. And this is why it mentioned
the fact that I'm sorry, I'm not able to do that,
what you're asking. But I can suggest you some
general fitness principles. So these are the general fitness principles that it suggested, the outcome and the thing that I want you to remember from Configuration right
here is that, first of all, if you want
something, for example, to your GPT program to directly get information
from your PDF, I don't want you to add to mention this in the Create part. I want you to mention this
in the configured tab, which is just a more direct way to communicate with your GPT. For example, if I go
right here and I say every time you, every time, let's say an answer no, every answer, Let's say
every answer with a hooray. Okay? So let's see if
it's going to work. How can I stay motivated
for daily exercise? Let's see if it's actually
going to end it with a hooray. But I think it's
going to, it's going to stay motivated for the daily exercise
can be challenging, blah, blah, blah,
blah. All right. And I think that now that
we've added this in the configured up, it's
actually going to work. This is like a more direct
way to program your GPT, which honestly
makes sense because this is just a starter
all starter way. The GPT builder is just a way
to start creating your GPT. And then in the configured
tap you're going to be just more
specific with it and the instructions here is pretty much exactly
where you're going to be putting all of
the information that you want the GPT to apply. It keeps on giving us
motivational exercises and motivational exercises and let's see if it's going
to end it with a. So again, to conclude, if you want to directly change the way that
your GPD acts. I wouldn't do it
from the Create tab. I would do it with
the configure tab. I would go into the
instructions and I would clearly instruct
the GPT to do something. For example, end every sentence, let's say with a UP. Okay, Let's see if it
does this actually. Okay, So tell me what you are capable of in
five sentences. So let's see.
Absolutely. Let's get pumped up about fitness UP. Okay, so it says I'm
going to go out to guide. Ok. And UP again. Up here, UP again, you can imagine how much
and how deeply you can tweak every single aspect of your GPT from the
instructions panel. This was just a demonstration of the fact that
whatever you instruct, this is exactly what
the RGP is going to do. So this, ladies and gentlemen, concludes our
configuration lesson. And this lesson we dived a
bit deeper into how to upload PDF files and train R GPT models based on those PDF files. And also how to deeply,
more specifically, again, configure the settings of RGPT
not from the create panel, but rather from the
configured panel, which is just a more direct way to communicate with your GPT. The final thing that
we're left to do now is to learn
how to share RGBT, which is something
extremely, extremely easy. And we're going to do this in the next lesson and the
final lesson of the scores. So thank you very
much and I'm going to see you in the next lesson.
10. Sharing our GPTs: So welcome now ladies and gentlemen to the final
lesson of this course, in which is going to be
a very short lesson. During which I'm going to
show you how to actually export your GPTs and how to share your GPTs to the public. All right, keep in mind that this course and this
lesson has been created exactly one week after the release of
custom GPTs by open AI. So the marketplace
in which we can list our DPTs isn't available
yet. All right. Chances are that if
you're seeing this course more than one or two months
after it has been created, there is possibly a
chance that there is a marketplace in which
you can directly list your PTs and
sell them there, generate revenue from there. Right? That being
said, at the moment, the best you can do
is that you can share your GPT with a code
to the public, right? You can copy the link
for your GPT and share it with your
friends or even, you know, sell it in
your own website. So let's go ahead and
actually let me demonstrate exactly how to share your GPT
after you have created it. So this right here again is
the interface of chat P four. And we're going to go
in the Explorer tab. All right, so in the Explorer
tab you can see that those are all of the GPT's that
we have worked with. And this is the
custom GPT that we have created, the
home fitness code. All right, so what we're
going to do here is that you can see that it
is locked at the moment. And we want to unlock it
in order to be able to list it to the public.
So we're going to edit. All right, and from edit you
can see that again we have the left side of
the screen which is the create and the
configured tab. And on the right side
of the screen we can again play with a GPT. Now how do we share it? We press on these
three dots right here. All right, we can delete a GPT. Next to those three dots, we have this Safe button. And if we click the Safe
button, you can set, we have all of those different, again, options to publish, either to only me, to publish
with people with a link, or to list it to the public. All right. As you can see, if you want to list
it to the public, you see that the
GPT may appear in the GPT store, which
is coming soon. So if you want to upload
this, for example, to listed in the GPT store
when it arrives, all right, and when it is possible to
upload your GPT and DPD store, you want to list it for the public if you want to
keep this GPT on a private. All right, setting for example, to share just with people
that you send a link over. It could be some
of your students, if you have a program, it
could be some of your friends. It could be some
of your teammates, some of your colleagues. Right? Then you only share
with people with a link. All right, then you can change, obviously some of
the features here. For example, the
theme of the GPT, some beta features of the GPT. Again, some data controls
and the bulk profile. You can either have or
not have your name. All right, and you can change, pronounce the name
again for the GPT. Nothing really special.
And once we're done, we press Confirm. All right. For example, right now
I saved only people with a link, right?
Press Confirm. And now if we go to
the Explore tab, you can see that right here, it used to be locked, but now you can see the
amount of people that have been inside this GPT
and have used this GPT. At the moment, obviously
it is zero, all right? But in the future, more and more people as
they use your GPT, you can see that this
is going to be 1234567. So pretty much this is
how you share your GPT. Again, if you want to
share to the public, you go here again. If you want to share
your GPT to the public, you go save public. All right. Confirm And now our GPT, as you can see, it is
completely public from here. If we change it, you could see from here down, right below, the name of the GPT it has published and
where it is published, so the home fitness coach is
published and to the public. All right, you can view
your GPT or you can copy the link of your GPT and
paste it whenever you want. In any chart you want to
share it with your friends, and this is pretty much
how we share ladies and gentlemen, our GPT's. This concludes the final lesson of the scores, ladies
and gentlemen, in which we discussed
about, again, how to share your GPT. I would like to thank you for completing this
lesson right here and I'm going to see
you in the thank you message of the course.
11. Thank you message: Ladies and gentlemen,
thank you very much for sticking
up at the end of this course and completing
this course to its fullest. I really, really hope that I've helped you with the
creation of the custom GPT. And during those
lessons in this course, I think we've analyzed
a lot of stuff. We've started by analyzing
what is open AI and how P three revolutionalized the way that we interact with
artificial intelligence. Then we moved by comparing
chart three with chart four, and then we dive deeper into the capabilities of
ChargPt four, right, and obviously the service
that it provides to us, which is the creation
of custom GPTs. Then we dived even deeper into
how to create custom GPTs, how to configure
those custom GPTs, how to import PDFs and
more information to our custom GPT model to tailor them to
appeal to the users. And finally, how to
share our custom PTs. How to upload them
in the marketplace, or how to keep them
on a private basis just by sharing their links. So again, thank
you very much for sticking up until the
end of this course. I'm honored that you
completed these lessons. If you enjoy this course and
feel like helping me out, you can leave a positive review. And if you want more
content from me, you can follow my
Youtube channel. Thank you very much, and I'm going to see
you in the next course.
LAMZ, Creative Internet Pioneer
