Transcripts
1. Welcome to Mastering ChatGPT: Want to use antipathy
for your business? Do you want to use JDBC
for your marketing, for your content creation, for anything that you do, that you put a lot
of time and effort, but maybe you don't
like as much and you would like stone when somebody is something that help you
do it quicker and easier. Well, you are in the right
place in this class. I'm going to tell
you how you can create your prompts or
the guidelines for chatty pithy to help you
in the best way so that you don't lose so much time going back and forth with it. And you have the best outcome, the best response that you
can use for your business, for your clients, for whatever
you are trying to use it. I know how frustrating it
is to try to speak the chat DPT until GPT doesn't
understand anything you say, you ask it for one thing, it responds with another thing. You try to do something, it doesn't do it for you. And you see all these experts online saying the
chat TBT is amazing, but you feel that
it's just a lost of time because
he's still eating. You are but you don't
understand how to talk to it. You don't understand how to
get the response you want it. Always you get this
generic content that you can not use
it for your business, for your marketing,
for anything really. My name is Alexandra. I am a digital marketing
expert and content creator, and also I'm in
love with ten dB d. I've been using Vali. They leave because it helps me a lot with everything that I have to do with my own business
and also for my clients. I know how frustrating it is to ask for something
or ask the kid to create content for your
business and being having these generic
response that you could find on Google very easily and is not how your
business speak. Also, I know how to work with it to have the
best outcome possible, to have that content
that feels and sounds and East like you like
you wrote it yourself. Tell deputy has been an
amazing tool for me for growing my business and for growing also the
business of my clients. It saved me a lot of time. It saves me a lot of headaches
and a lot of research. And I wanted to help you. I've created this
class because I want you to learn how
to create prompts or guidelines to ask Chad
DPT whatever you wanted, whatever you have in that mind, whatever it would be helpful
for your business and get the response easily
without loosing time. This class is very short, it's very to the point and you will have a formula that you can apply to anything you want to do with the
help of Chad GPT. By the end of this class, I assure you you
will see chat UPS, your personal
assistant, and you will be in love with it as I am. This class is for
total beginners. I don't need you to have
even an account on chat DBT. We will do everything here. You will learn everything, all the basics here. So you can kick
start your journey with these amazing AI tool. This class is not for
people that is looking for advanced techniques
in prompt engineering. We're not going to talk about anything about machine learning, ai or prompt engineering,
advanced techniques. So if you are looking
for that type of class, this is not the place. This is a place
if you're looking for something very easily to understand a formula that you
can apply in every prompt, you create a way to use GDP
t for any endeavor you want. What do you need for
this class is basically a computer and a connection
to the Internet. If you can get
connected to chat DPT, then you are fine
and okay, good. Suggests you as well to have a word processor like
Word or Pages or notion, or whatever you use to storage any type of written
documentation because it will be easier for you
to take notes and to have the prompts that has
been working for you there. And that way it's easier
for you to experiment. Let's say no more. I'm Mrs, start creating amazing prompts and
guidelines that we'll get to DVD to work for you
easily without headaches, without hustle,
without any lost time. I'm sure you will enjoy
this class and I see you inside in the
next module command. Let's go during the journey, it's going to be fun.
2. Class project : Who I'm so happy that they've
decided to take this class. Let's talk about the
class project and how everything's going to be
structured in this class. So you can get the
most out of it. So the class project that we're going to have is just basically using chat DPT with all the instructions
that we're going to see, all the key elements and the formula that I'm
going to teach you with the prompting engineering is very important that
you take action, Go to tell GBD, experiment with everything
into action after the modules. So that way you're going
to learn how to use it, you're going to get more
familiarized with it. And mainly due, are
going to learn how is the best way for you personally to interact
with chat GPT. I really feel that
this is very personal. This is something that I
can teach you the formula. I can teach you how to do it, but I can not be you. So it's very important that
you take action every time. Basically what we're going to do here is that I'm going to teach you the key elements
of prompt engineering. Everything that you need
to keep in mind so you can kind of like replicate
these all the time. And you can create your
own prompts and you can sort of generate
your own thing. Because I feel that's very empowering and that's
something that you need. And also I'm going to give you a formula which you can
follow step-by-step. After some of the
modules, I will tell, I will tell you exactly
what to do and I will guide you to create
your first prompt. But of course, I always, always, always encourage you to
experiment with Chad DBT. So please don't just wait till I tell you that you go
to write something. Just go there, write something and just sort of like try to get curious and creative
with it because it's going to help you
a lot in this journey. At the end of the class, I hope that you submit
your project and you show me exactly what was
the prompt, you create it. What was the answer that
I gave you and what was the iteration that you did
with it and the final product? I'm sure it's going to be amazing if you have
any questions, if you have any hiccups, if there is something
that you are curious about and
you are not sure, you can always ask me in
the comments of the class, I am always there
keeping an eye on you, seeing what you're doing, seeing if you have
any questions, I can help you personally and also please
leave your projects. So of course, I want
to see that prompt. I want to see
everything that you do. Thank you very much again for taking this class and
let's say No more. Let's start with the first
module where we're going to see what exactly
is contributing. See you there.
3. What is ChatGPT: Okay, let's go for module one. I'm so happy that
you're here and dry now we're going to talk
about what is Chet GPT. I want to give you
an introduction of what is this technology
that we're going to use. So you understand what you
are using and you don't use do things without even
knowing what they are. And that's the goal today. So let's go ahead. So each activity
has gained a lot of popularity during the last year. But what is exactly chat DBT. Dbt is stands for generative pretrain
transformer is basically a natural language processor
program created by OpenAI that understands what we say and can give us answers. Openai is a company that
has been working with machine-learning models and AAA to give out different programs. Gbd is one of the biggest one, and it's basically
changing the face of AI and how it given to
the people in the world. Chelsea PTs, the chat bot interface with which you
can interact is when you open chat GPT in your web browser and
talk to eat them. Yep. You know what I mean? Less worse is basically like having a conversation
with someone that knows a lot about different things and can give you a lot of different answers. But you need to be
very careful with the questions that you
asked this person. Robot. But because in the questions is the quality of the
conversation that you have. If you asked very
crappy questions, you will have very
crappy conversations. In the words of
charity, beauty itself, chat diabetes, like a
very super smart friend, then can have different
conversation with different people about
very different topics. But as you're smart friends, you need to know what
to ask them if you don't want to have well,
a crappy conversation. So what is the most important
thing about charging pithy is the type of
questions that you give. That questions are
called problems. That's why we have
this class today because I'm going
to teach you how to make the questions in a
better way so you have a very good conversation
with it and you can have the answers that you need it
to improve your word to be more productive or to
have a great time. Okay, so now that you
know what is steady PT and what is the
GPD technology, we're going to create your own account so you can
start using this technology. Basically what you
need to do is go to this URL right here below in my screen that
is Chad open ai.com. Then you're going to find
with this screen right here, welcome to login with your
OpenAI account to continue. You don't have an
open AAA account. Well, if you have
one, please use it. But if you don't have one, Let's create a new one. You need to click on, Sign up. After this, you
have three options. You can use an e-mail address. You can continue with Google or you can continue with
Microsoft account. Of course, you select
whatever is best for you. I do prefer to use Google, so I'm going to logout
login with Google. Okay, So after you login with Google or with the email
address you prefer, or with your Microsoft account, then it's going to
ask you to tell it your name and birth date. Of course, you can
use an alias always, but let's use our name. So I'm going to put my name. And then I'm going
to put my birthday. And then you need to
click on Continue. Then it will verify
your phone number. You use your phone, you're correct phone number. And you will click
on send a code. Then it will send a
code to your phone. And you can enter that code
in this space right here. Once you enter the code,
everything is done. You have access
now to check GPT, of course you have access
to the free version. There is an upgrade or a plus version that you
need to pay for it. If you want to access
to the plus version, you need to click on
Upgrade two plus, and then click on the
plan that will select, that will be Chad DBT plus
right now is $20 per month. If you want to have to it, then it would be fine if not
for this specific class and actually for everything
that I've done which LGBT the free version
has been okay, so you don't need
to buy anything for this specific class or even
for your day-to-day actions. If you find that you get
limited by the free account, then you can always
purchase the plus one. Well, that's it. Give us a little bit of examples, capabilities,
and limitations. Of course, we're going to talk about everything
here in this class. So you will have all the information and
even more with prototypes, so no need to read them. But if you want, go ahead and you can start typing
anything e.g. you can say, hi, your child to be
able to say hi back. So now it's time for
you to take action, go and create your own account. Let's go and start getting
familiar with Chad GPT. Let's go to the next module
where we're going to talk about all the
capabilities such as TPP.
4. What can ChatGPT do for you?: Welcome to the second module. Here we're going
to talk about what the capabilities to LGBT has
and what it can do for you. So let's talk to the source list as Chad DBT itself,
what it can do. So let's great. Okay, So this is a lot of text. If you want to read it, please hit pause the video, read it through, and then
hit Play to come back to us. But I'm going to give you
a TLD or version of it. So what content
typically do for you? First of all, it can talk
to respond like a human in different languages
in several topics. This is one of the most
impressive parts of these AI. The capability that
it has to talk and interact with you in
different, different topics. Really, you can ask it about such a variety of topics
that is impressive, how much he learned and how
much it can discuss with you. And second, how friendly it is. It feels like you're
talking to a teacher, to a friend, to a mentor. It's amazing how
easy it is to use and how friendly it is to
discuss different things. That's why it's very
powerful and it will help you a lot in your
day-to-day actions. But maybe you're thinking, Well, Allie, I actually don't
do things in English. I do things in Spanish, in German, in French, in Arabic. So then I cannot use IGBT. Well, you are drunk. Actually check GPT can talk
in more than 20 languages, including traditional
and simplified Chinese, Noriega, and Bengali and more. It's impressive, but you
need to keep in mind that the vast majority of material
that it was preteen, pretrained with,
it was in English. So the accuracy is lower
in other languages. So just keep a night. Also siesta Hannah
and the continue and language heck at the
end, indifference, Yeah. Because animals
differently as I handed or the formally that ecosystem
poco mas complexes and quantize to DNS because there's
more and more specific, our specific 0 and quanta
kettle boiling Gua Hikida sketch LGBT who say no lower IN. So you need to keep in
mind languages, french, German, other languages that
has different complexities. Be very careful and be very
specific on how you want the language to be in the
response that gives you. The second thing that
it can do for you is learning based on data
that you give it, meaning that you can train it. We're going to talk more about the training phase
in other modules. So let's keep it in mind and we will talk deeper in that module. The third thing that
it can do for you is that it can be your
personal assistant. It can help you do things like time management,
task management, and even you can use it to
keep track of your thoughts or brainstorm or kind of like bounce ideas
back and forth. That is my favorite
feature of JEPD. Basically I'm having
conversations with them to really like is my personal system used to keep track of
what I need to do? What are my ideas? Just to have a voice like
bouncing back my ideas. So that is amazing. Please be mindful that since we are going to talk
about the free version of charges and we're not
going to dive deep into the coding
part of chat GPT. You are not going to
be able to create apps to ask Chad GPT-2, make a calendar invitation
for you or things like this. These needs some
coding knowledge. We're not going to do that, but still is something
that is a capability, but it's paid capability
and you need to have more knowledge into
coding and data management. So we're not going
to go into there. The first thing that
it can do for you is that it can be
a creative mind. Be aware that I have
here a little star, because we need to be
very careful with it. The GDP can be creative. Of course, it can
never replace a human and our tree, our
creative power. So you can use it
to bounce ideas, to ask for other ideas, or to brainstorm,
but don't forget, check GPT is a
machine-learning model. Let's repeat it again. So GPP is a
machine-learning model. Please don't go into the rabbit hole or
reduce asking things. Copy and paste
into your website, or copy and paste
into your book, or copy and paste without
even really through, without even putting your input, your input and
your uniqueness is very important for
the work you do. So dP, dy is more
of an assistant rather than it's going to do with things for you and that. So this is very, very important. Let's not lose sight of it. We can use it for good, but we can use it
as an assistant. But let's not use it to replace us and replace our
uniqueness in our business, in our creative endeavors and everything that we're
going to use it for.
5. Limits & Limitations of ChatGPT: This is a very important module, so please do not escape it. We're going to talk about the
limits and limitations of Chechnya petty and
how my slice it. Be cautious because
this is a very, very important information
that you need to be aware of. I'm almost sure that you've
seen in the news that e.g. things like Chelsea
beauty can be sentenced. It is like a human. It is very evil and
things like that. We need to be very aware. Chat DBT is
machine-learning module is a language processor. It doesn't have a consciousness, doesn't have human ability to make a conscious
decision about something. What that means is
that the ethical and the biases of charge come from the people
that created it. And even if people that
created it doesn't want it to have certain
biases are, certain. Ethics is very difficult for
GPD to not end up with it. If the data that you
input through it has some biases or has some ethical so that
we don't want it to. E.g. IGBT can be very negative
or it can give bad advice, stores hurting people
or be harmful. Or actually it can be a very, very bad advisor towards your mental health because it doesn't have the ethics of us, humans that we learn
from very young age. What is good, what is bad? We learned how to make actual decisions
not to harm people, are not to be evil
towards people. This is a very important
discussion towards AI, and here's how it
gets strain and what is the data that you can
input to towards it? And what are the
questions that you can ask for not to be, for it not to be biased, or for it not to have ethical challenges into
the answers that it gives. If you are interested
in this part of the AI, I recommend you to hear the
episode of the last legs Friedman podcasts that he
talks with the CEO of Open AI. They go really
deep into ethical, into how difficult
it is to align your GPT thoughts and with
what is good and what is bad. And that is basically also a human question of how
as humanity as a whole, as an entire
community of peoples and entire planet
can define what is good and what is bad and be very cohesive around everyone, what that alignment
is going to be. So that is a very interesting discussion
that I suggest you to here is going to be down below if you want
to click on it. But basically what I'm saying here is that GPT doesn't have the capability to make
a conscious decision of not to give you information
that can be harmful. So it's very important that
if you are asking e.g. anything that is
around mental health, anything that is around maybe ethical issues or
philosophical issues. Please be very mindful and very thoughtful and read
through the answers, but take them as a grain of
salt rather than as a whole. You will see also as
well in the internet that has something
as a jail break. And you can make it same thing. You can make it very harmful. You can make it very negative, or you can make it
be very blasphemous, is something that it can happen. I will encourage you
not to try to do this because this is not why the
technology was created. This is an amazing
technology that can help us do amazing things. But if we go to the round of how bad I can make check UP TB, well, you're going to find it. So let's try to focus on the positive and on the help
that it can be for us. And maybe let's hear more about the experts and let's
try to understand how we can align it with good ethics and owned
by us information. So we avoid this type of stuff and other parts of the
limitations that is very, very important that
you keep in mind is the DPT was trained with information from
2021 and before. So if you are asking for news
or any or any fact check, this is something that
tells you cannot do. Do not use it for
news information. Do not use it to
fact check something because it's not
accurate on that. That means that if
you are working e.g. for a newspaper or anything that is from 2021 towards now, you need to be very careful
and be aware and read through and act like an
editor into these answers. Use put a lot of intention
into really makes sure that everything that GPT is giving you and everything that
is generating towards, towards the question that you
asked is actually correct. Fact checked everything. If you are talking
about hard fats. And also really true always. This is something that I'm going to tell you all the time, read through the answers. Don't copy and
paste and that sit within the limits and
limitations of children pithy, you need to be aware that
Teddy BTS owned by a company up as sort of like mixed
between a private and an 0, 0 and G Company. It's very, very important
that you keep it in mind because there is an
issue of privacy. I've seen a lot of people using GPT for clients or
for their work. And they normally up
the upload everything from their from their company to charge you asking
for questions. E.g. I've seen developers use uploading all the code from
production to chat DBT, just asking to improve in above or make a correction or
something like this. And this is something that
shouldn't be done. Why? Because we don't, we
don't exactly know how Open AI is storage
that information. If somebody else use each LGBT as well can
access that information. And if you are sharing anything
that is confidential or the children leave your
office or your computer, then that is a problem. I always ask one thing when I'm uploading something
to charge UP team, where I upload these to read it. When I upload these to
any public interface, if the answer is no, then I need to take
away everything that is confidential and everything
that I should have put e.g. please do not share your
personal information, something like your knee. That's fine. Your last name. That would be fine as
well because you can find it in public
records, but e.g. your social security number, that's number, bank account, anything that is
private information that somebody can
use it to harm you, you shouldn't upload it there. It's very, very important
that you keep this in mind, mainly if you are doing things like e-mails or e-mail campaigns which LGBT sometimes we share confidential
information through e-mails, and it's very important that you don't share it with Chad, DPT, I would say even be
more aware of these rather than just thinking about
making chair deputy evil, use it to be helpful for you. Use it for good, and use it carefully. And this is everything
for this module, I think you now know
what are the limitations and the limits you should
be aware of which LGBT. Let's go to the juicy
part of this class. Let's go to the prompt
engineering module.
6. What is Prompt Engineering: Welcome to module four. Here we're going to talk about what is prompt engineering. So first of all,
what are prompts? Prompts are basically
the instructions or wherever you write
to chat UP team. It can go as simple as
what is one plus one? Or how much is one plus one
to a huge tax like this one? Where you want a simple answer like one plus one equals two, then there is no much to think
about the prompt, right? It's very simple. Although be mindful of all
mathematical questions, because it seems
like Chad derivative gets very confused very easily with mathematical questions and it can give you bad answers. But here the keyword
is confusion. What we want to stay away or keep jack chat GBD
away is from getting confused when you
are trying to use chatty pity for your business or for your creative endeavors, creating e-mail campaigns or brainstorm new ideas
for your book, or bounce ideas back and
forth for your business. Then the prompt
start to get very, very complex and you start to increment the amount of
things you tell chat, DPD, aka the training,
then it's very, very important
that you keep a is structure on our way
to talk to LGBT. So we started from confusion and be very
specific and clear. So we have the best outcome. So what is prompt
engineering world is shaping the
instructions that we give to chat DBT in order to get the desired responses in less. The most important thing is
to keep away from confusion. So you get the answer that you needed in the format
that you needed it. But here let's talk
advanced for us. Second, as I mentioned, there is a more
advanced way to use chat GPD or the DBT
technology with OpenAI, they have an API that
you can use to create your own queries and you do
this through lines of code. So it's important to have
some coding knowledge. You can use it e.g. to process large sets of data or create your own programs
in that context from engineering is more
advanced and you need more technical knowledge into what AI and machine learning. If you want to dig deeper
into these topics, I will leave at the
end of this class some links and some
spaces where you can find more information about it and continue
your learning journey. But for the topic of this class, a specific we're
going to keep with the basic techniques
of prompt engineering. And we're not going
to use the API, but instead the chat bot
interface that really is enough for your business and you can do amazing things with it. So now that you know what
is prompt engineering and the most important
thing which is keeping Chen GPT away from confusion. I encourage you to take action. Go to chat GPP and experiment. Go and gray, some
instructions go on, right? Some questions that
you have in your mind. I know everyone has that question that wants
to ask to tell dP, dy. So this is your
opportunity to go and write whatever you want to experiment with it and try to get the feeling of
how it answers two, you get the feeling
of how would you like to use it in the future. Because in the next module, we're going to dive deeper
into the prompting formula.
7. Best Practices for Prompting: Hey, welcome to module five. We're going to talk
about prompting the key elements you need to keep in mind for creating
amazing, amazing prompts. It's very, very
important this class. So pay attention because it
will set you up for success. So let's start right now. The first key element
of prompt thin, which is not surprising, being clear and specific, we need to be very, very clear on how we
ask the question, how we guide chat GPT on
giving us the answer, the outcome that we want it. Because if not, it's going
to get very confused. If your prompt is very
confusing, very messy. It doesn't have an action verb. It is not clear on what you
want the outcome to be, then probably you're
going to have a very bad outcome and you're
going to lose time just trying to get DPV to
get to the clarity that you want it to be
clear and specific is very, very important that you, yourself are very clear on what you want the outcome
to v, Please. I'm going to repeat this
again because it's, it's very, very important. We need to have a clear
vision of what we want out of GPP for LGBT to be able
to get to that point. If we don't know, because sometimes we don't
know exactly what we want. Probably we want some guidance. We want something
to help us in maybe creating ideas or maybe grade. But we really don't have a clear idea of how
the outcome should be. We can use shall GPT-2 guide
us to get that clarity. But if you are not clear about
the outcome that you want, then chat typically won't
be clear about it either. The second key element
is providing contexts. This is very, very
important and this is basically consider as
the training phase or the training data. You need context, that
you need to let it know. All the information that
it needs to know for me. Here's a pro tip, is thinking DBT as a
person that knows a little about a lot of things and probably has the
right answer for you, but has no idea what you're
talking about in the moment. If you don't give it a context, if you give me permission to start to generate
ideas from them, is going to be amazing. Conversation is going
to be very hard to get to the point that you want
to be providing contexts. I'm also talking about
providing examples. You can e.g. say, Hey, I want you to create
an Instagram post, and this is how I
grade Instagram posts. Look like. You give me an example
of something that you wrote before or Instagram. That way it has the context
on what exactly you want to have and it can
give you that history. Here. You need to
be very specific on the limits that it has heard, the restriction that it has
in the outcome that you want. E.g. if your business
only talks to women, then you need to be very
specific and tell it, okay, you are generating e.g. an Instagram post for women. So that way it knows what the target area is and it
knows that limitation. Also be sarcastic or funny or inspiration or how you
want this outcome to be. Those are restrictions you
need to tell charge if it is not going to assume
it. Are there limits? E.g. how long you
want the outcome to be doing a blog post that
is optimized for SEO, probably you want it to be
longer than 300 characters. You need to tell these because it's not going
to be by itself. Unless you are very, very clear, there is some CNS blog post and you already seen it around, that it needs to be as well. If you want to include
any ethical guidelines, you can do it briefly
about voiding, avoiding certain type language
or avoiding type of IBS. You need to think about a five-year-old and
tell it what are the limits? So here you need to tell it the next key element and
he's including keywords. Keywords are very important
because it will sell DVDs attention and probably will use it as well in the
outcome that we generate. So it will more
your language and the type of answer
that you're expecting. You need to use keywords
and key phrases that are important for your business
or for your clients, or for whatever
you're asking me. Do you think through
those wealth, industry jargon or any type of language or a specific worst of the user for you or that you
want to use in the outcome. It gets even more
relevant if e.g. you are generating a blog post for pharmaceutical
industry company, then going to be
very specific on that jargon or that link
we're going to use, rather than writing
something e.g. a. High school or university
degree, it's very, very difficult how the
terminology is going to need to be very specific. Here as well. I like
to include in keywords what is the type of audience
that I'm going to talk to? Who knows exactly how to create. Answered that. The other element, the other key element
that you need to keep in mind is verified and
cross check answers. This is very related with the previous module
where I talk about the limitations and
as I told you before, Chelsea was pre-trained data
from 2021 and before that. So anything that is from 21 to the state doesn't
know who he is. Very, very important that you cross reference everything that is generating and verified
all vowels that is getting. Again, don't just copy
paste and that seat, but just read it
through and make sure everything is cohesive and everything we eat should be if you're
including hard facts or if you're including something that should be cross-referenced, then cross-referencing, use Google makes sure
that everything is okay, then you can use other
key element is to specify a format that you
want your question to be related with
setting the limits. But here I'm talking about
how the answer should be, what is the format
that it should take? How it's kind of
like the packaging, the package in which
should be dropped. E.g. if you want something
to be a bullet point, or if you want something
to be a table, or maybe you want to scrape
with every single element. Or maybe you want a blog post without even knowing what
is the title or anything, you just want the texts
that we need to be very, very specific to how the format. You want, some things it
can be in fear from e.g. if I ask you generate
ideas for blog posts, then probably it's going
to give me a number list. But sometimes it's
not that easy. So you need to be clear. If you have a very, very, very concise and clear idea on what outcome
you want it to be. It used to be mindful
of communicating that, to check GPP so it knows, okay, these ideas, I should
generate these like these, unlike these formats, you could also include
something, e.g. all the answers you give me, nice generate title,
description and keywords, squares that then typically
knows at all the answers that it's going
to give you after that are going to
be bad for her. Title, description, keywords,
title description keywords, unless you ask for
something different. Afterwards. The last key element, and this is actually
a productive for you from everything
that I've experienced with Ted UP is that
you can control the detailed level that
you want in your answer. Sometimes we went very short
answers like a like e.g. what is one plus one would want to know if that
is true and that's it. We don't want more
detailed explanation. We don't want conflicts, we don't want more information, but sometimes we want a long
detailed answer about it. Sometimes we want something
that is more process, something that is more for, that is actually bigger, longer, that has more context to it. To make sure that check GPP, create the detail, the level
of detail that you want. You can use the
prompt to do that. And basically, what you can
do it is that if you want a shirt for your problem
should be shirt. Gpt knows that. Okay. This is exactly
what she wants to know and this is
exactly what my answer. And that's it. If you want the detail of
your answer to me very low, you need to create
our longer and ask open questions about Teddy. Begin with longer
and we'll explain more in detail and more in-depth anything
that you're asked for. Okay. And now it's time
to take action. Lists, everything that
you've been thinking about, asking chat GPT for this, everything in your
business can be automatized or can
be helped by chance. E.g. if you run a creative
business and you sell art, you can ask your DPT
to help you with your social media content creation, with the scripts
for your videos, with ideas on what type of things to pose
with a calendar, what are the best timing, etc. Think about how you can
help your business. So that way, you have some ideas for the next
module where we're going to use the prompt formula to create the best prompt and get
you the best results.
8. The Role: We are in module six here we're going to talk about
the prompt formula. I'm going to cut this module in different
modules so it will be easier for you to digest and understand each part
of the formula. Of course, at the end,
we're going to have a summary here in Module six. We're going to talk about the first element of
this prompt formula. Let's go. The first element of the prompt formula is the
role you need to tell it how to act like when E is generating the
answers for you. This will make it very
easily for GPD to understand what type of knowledge or what is the
direction they asked for. It should go. Not the same to act like a
digital marketing expert rather than to act like
a video content creator. Because it will give
you a different set of skills or knowledge that you
need to apply to the answer. So this is a very crucial step. Please never skip it
unless it's a very, very short from that u1. A very open and
very general idea. Here I have some examples
of different roles. I asked Chad DPT, to act like I only gave you the prompt of act
like dot-dot-dot. It will tell me
exactly what it can do for me when it's
acting in that road. This is a pro tip for you. If you have no idea
how to use GPT, you sell it at light up. And then it will give you a list of things that it can do
acting like that, e.g. act like a personal trainer
and it will tell you exactly what it can do for
you as a personal training, trainer or acts like a
digital marketing expert. And it will tell you
exactly what it can do for you as a digital
marketing expert. If something that is not
listed there for it, hey, as a digital
marketing expert, can you help me in
these nice on that? And it will tell you yes or no. This is one of the
best things from Chucky D that is very,
very conversational. So use ask things and it
will give you answers. Okay, so let's look
at the example. So here, the first one, it was the very basic one app like at digital
marketing expert. Basically it tells
me Sure thing as a detail mass marketing
expert, I can provide. You can read it through the
second prompt that I give it. It was more specific
and more to the point. You see that you can be more specific within the role
that you give chat dP, dy. I ask it at like a UX experts we've experienced in B2B and
the e-commerce industry. That way what I'm doing
is telling it, okay? The answer is that you
need to generate for me, needs to be guided or needs
to be pointed towards the e-commerce industry
and also a B2B business. That way is more easily for
me to get the answers that I need because that is more
specific towards mind needs. The next prompt that
I gave it was at like a personal trainer
with ten years of experience in weight lifting. One thing here is that
I want you to notice is that I made an error
here in my grammar. I am not a native speaker. So I actually do a lot of
errors when I'm talking to GPD. And it can catch that I'm doing an error that is a
typo that maybe I misspelled something and
it will give me actually the answers that I needed without being my
perfect English. And as you can see here
in my prompt, also, I noticed how much years
of experience why I needed to tell charge EBT here is
that I want a senior answer. I don't want every light as her Barry on the surface
surface and swear I want something that dig
deeper and that I can feel that do know a lot
about this topic. The next prompt that I gave
it was the most complex one. I think it's the most
interesting one because I, I told Chuck DBT, act like a social
media content creator with experience in short format videos
for creative is small businesses like
artists selling paintings, selling their paintings online. What am I telling IGBT here? Well, I need somebody
that knows about social media content and also that knows a
lot about videos. So it can infer
that probably I'm going to ask for something
around the videos. Then I use guided tours. What is going to be
my a specific need? My a specific need is going to be towards the small businesses, creative businesses
that I'm selling my paintings that way you
can give me very, very, very, very specific answers to the questions
that I have after. So let's summarize. This is given the role to chat. Dbt is the first part of the prompt formula and
it tells Teddy PT, how to add and how to
guide your response. You can be as thorough
and as detail on the role that
you want it to have or as light on the role, I want you to act like x, x or I want you to act like x, x with this amount
of experience. And these are specific
specialization in these specific industry. And that way it will be very, very focused, all the answers
that it will give you. And if you have no idea
what it can do for you, you can just ask it at XXX and it will give you a list of
things that it can do for you. Let's go to the next module
where we're going to see the next element of
our prompt formula. And we're going to talk
about the changing face.
9. Training ChatGPT: Now let's talk about the second element of
the prompt formula, which is contexts, aka the training, as I mentioned before in
the key elements that context will give
all the information that needs to generate, everything that you
want to generate. Basically what you need
to tell it is paid. Here is a context
about my business. Here's the context
about my client. Here's the context about what
I want you to answer me, or here is the context
in which you need to locate yourself to
give me this answer. The role we told it, how to, how to act like, right? In which position
to put itself to give us the answer and
with the contexts we're going to create the ambient in which it needs to
give us the answer, but exactly what it
means to give contexts. Well, these are the
six points that you need to keep in mind when you are
grading the contexts. Or basically when you
are training your model. What do I mean with
training and why? I keep telling you that
it's very important. I mean, that is something that has some pre-trained, right? That is the PI DPT that
has some pre-trained face, so it knows things even
before you talk to it. But for AI is very,
very important, the training phase, chat EBD already know something but
he doesn't know everything. You need to give it all the information
that doesn't know yet. That's why at the
beginning of this class, I told you GBD is like talking
with someone that knows a lot about different
things but knows nothing about whatever
you are talking about. So it's very important
that you kind of like give all that
information so it can sort through all
the data that it has already and find the right data to give you all the answers. E.g. if you are asked to create an Instagram post is
very important that you give the training
into what for us specifically is a
good Instagram post. If not, you're going to have a very generic answer
because it's going to take whatever it has already in storage is basically
like googling. Give me an Instagram
post, copy paste posted, and thinking that is going
to tell anything special about you or your
brand or anything that you're doing is
very, very important. This training phase or
this training material, because it will tell it exactly how it needs to generate
these answer it, we'll give it the
ambient, the context, everything that it needs to
know from the background, so it generates the
best outcome possible. And that way you won't need
to do so many changes. Again, you need to read
it, verified, etc. But you won't need to
do as many changes as maybe taking use a
generic post and you have to work through it a lot different points that you need to provide
in the training. Our first, you need the context of the
brand and the business, the person that
it will work for. Basically you're
telling it, okay, you're working for me as a
client and I am Alejandro. I talk about marketing, I talk about content creation. I create Skillshare classes, I create digital content, etc. Today I do need to tell
everything about the brand. If you are working for
a client with Chad GPT, then you need to tell
it about a client. Who are they? What is their
mission, what is their goal? What they talk about, what it is the thing
that makes them unique. Here is very important
that you keep all these branding concepts and all these brand things
until it to charge it. Mainly if you're using it to create marketing content, e.g. or anything related
to the Internet. You know that uniqueness
is very important. So please give it context about the client or the brand or the
person that will work for. The second part is for
who is the content that is creating for basically what is
the target audience? That's basically how we
call it in marketing. Who am I talking to? Am I talking to women? 15-20? Am I talking
to men, 50-60? Am I talking to
experts in the field? Am I talking to a five-year-old? Who am I talking? You need to tell Chaddi, PT, who is it talking to? E.g. one thing that
I love to do and I do every time is
asked Chuck GPP, Hey, can you explain
me this concept? You're talking to
a five-year-old, I'm selling exactly what is the target audience
of that content, or how it should
be put together, or how it should be structured so that person understands it. The third part is giving any relevant background
information. E.g. you can give it a
background information that hey, I'm Alexandra, this
is my business. I'm talking to people
that doesn't know about Chet GPT and the bakery for information is that
I work in marketing. I know a lot about marketing. I've done classes before. I'm not a native English
speaker, you know, like just give it some
details so it can kind of put them uniqueness of the
thing or of the person. Or the brand into whatever
answer is creating the other training
data that you can give it is the goal
of the outcome. E.g. if you ask it, okay, please add like
personal training with ten years of experience. And my goal is to
be able to lift 20 kilos in two months because I haven't
lived anyways at all. So that way it knows what is the goal of anything
that you're asking you. And it's kind of like craft
everything towards that goal. It kind of like give
it a direction, have very, very
strict direction. Also, other thing that
you should include in your training material is
examples of scenarios. I've, I've told you
this before, e.g. use copy and paste, whatever is a good
Instagram post for you, or give it an scenario. Okay, So I already created
an Instagram post in these, in these contexts and
it was very successful. So I want you to create
something similar. The last thing,
and this is again, a pro tip for you, is as if it understood, basically just put at
the end of your brand. Do you understand everything? Is everything. Okay? Is everything correct? Do you need more
information from me? Do you know what it means? Everything this way. What Chad DBT will generate is a summary of all the
training that you gave it. You can read the summary and
see if everything is clear, if it got everything right. Or maybe you need to tell it. Actually know you
got this wrong. This is the right information. You should process these,
these, and these way. My prototypes here are
basically these ones. First, pretend like you're
talking to someone that has no idea who are
you, who is the client? Or the reason is that
you're working with. Think about someone that knows a lot about a lot of things but has no idea in the world. You are your business
or anything like this, use treated like that. The second one is use bullet points on
important information. So just verbally, identity, target, audience,
business goals. You can use these
list items list, and just fill it through to
give all the training data. You don't need to use
all of them because they are not relevant
in all cases, e.g. if you're asking a very
specific question like, Hey, I need to explain
it to a five-year-old. Why the why the sky is blue. You don't need to
give more contexts. You don't need to
give a gold jewelry, need to give anything
else because it's very specific what you need. But if you are creating e.g. our social media calendar with all the content for your social
media is very important. You take time to train
it very thoroughly. So that way the outcome will be better and better every time. The last one, and this is something that I
found that I loved. You can give chat GPT links. So it will scan through
the page that you gave it. You can kind of like
copy paste the URL. Maybe you have a URL where you have the ultimate guide to create a social
media calendar. You can copy and paste
and just give it to it. It will crawl through all
the page, learn from it. It will give you a
summary as well, what it learned from
that page and then it will apply it to everything
that you ask it. This will save you a
lot of time because you don't need to create
the material. You can just have your reference URLs and just give
it to chat GPD, and it will be
faster and faster. Okay, so this is all
for this second part. We have now two elements
for our prompt formula. We have the roll out at liked the dot and we have the contexts
or the training that is, Hey, this is what
you need to know to create everything that
I want you to create. So let's go to the next
module where we're going to talk about the third element
of this prompt formula.
10. Action Verb: We are almost in the middle
of the formula for prompting. The next one is very
interesting and it's the one that is going
to generate everything. And it is having a clear action, aka having an action verb, you need to tell directly to
charge UP exactly what is the action that you want
it to generate or to do. E.g. create,
generate, brainstorm, do great, re-grade,
roofing, bounds me ideas. Make a table, create a script. You need to tell
it exactly what is the action that you
want it to take. If not, it's going
to have a lot of information and
probably is going to give you some answer, but then it won't be the actual answer that
you were looking for. Or maybe you are very lucky and it is the
direction you are looking for. It's so, so, so
important that you don't forget to add the action
that you want it to take. I know it's very easy
to forget with between you're creating all the
training and act like XXX, then you start forgetting exactly the action that
you wanted to take. It's good that you
do the training is good that you give it a role. But then you need
to clarify what is exactly the action you
want GPD to perform. Alright, so let's young to each IGBT and let's see
some examples of this. Okay, So here we are right
now into chat DBT and we're going to try to
do some things with it. So first of all, we're going to ask
you to act like a digital marketing experts, a specialized,
specialized in SEO. I loved that when I
use it all the time. Then I'm going to tell it, I have a business
blog about, right? And then I'm going to
give it the action. This is the part that is
very, very important. I'm going to do to generate, That's my action verb. Generate a list of five persuasive and SEO optimized blog title. Okay, Let's go
ahead click Enter. And then it's going to start
generating some of them. I love that it always gives you, gives you serve like a summary of what is
this going to be doing. Because that way you can kind of see at the beginning
if it's doing exactly the right
thing and you can stop generating if it's
not doing the right thing. Because sometimes you
mess up or do ask for something that it was
not really the one. Or maybe it just
misunderstands you. And you can read that
very first summary, which is this one and see, okay, It's
understanding something else so I can kind of stop. Just click here well now so
three regenerate response, but I can click here and stop generating and
then ask it again, We hate you didn't
misunderstand me or actually what I wanted
you to do was XYZ. And I can see here you
see it showed me a list. I love that since I asked you to be digital marketing experts, specialists in SEO,
and I asked for the lease to be persuasive and SEO optimized block titles. It gives me kind of like an
explanation of why the title is persuasive so I can sort of understand what is he's
thought of process. And then it gives me some of the keywords that are relevant
to that specific title. So basically I can serve like, understand why this
is coming from. Is it coming to me and
how should I write the blog or how should I
ask it to gray the blog? One thing that I forgot
here and that actually, he should have asked me if I if I instructed
him to ask me that, you should have asked me
what was my target audience. So let's just do
another one right here. Okay. Perfect. You generate, I always loved to say
Could you please? Because that makes
me feel like I'm actually talking
to an assistant. It just makes it
more humane for me. I don't know. Maybe for
you is not that good. You generate a block bows
for the first title. Keep in mind this give
it some contexts. Keep in mind that
my target audience, our art lovers, one from it. And then we click either one
thing that I always tell you and I'm going to take
this opportunity to tell you once more. Please do not just copy paste. I'm post this. If you're going to do
this for your blog, you actually need
to read it through. Understand? Is this the kind of boys
that I use for my blog? Is this something
that I would use and also include the
things that makes you you that makes you are professional and makes
you the marks there. Because I don't want you to
be lost within the chatbot. But this is amazing. I mean, this is a lot of
work that saved from us. And see, I only had
to give it the row, a little bit of context. I didn't give too much. I gave a little bit of context. Again. I gave a little
bit more context when I ask it the second action. And then I asked you to generate they titles and it
did amazing work. One pro tip that I tell you is that within the
same conversation, you should keep the same type of information so you
don't have to give more context or change the context in one
moment to the next one. And also, you can ask for different actions in
different problems and just assume that it already has the training so
you don't have to train it and you don't
have to do this formula, the exact formula all the time, but it's just sort of like drags over information
and that's it. So let's go to the next
module where we're going to talk about the next
part of the formula.
11. Set your Expectations: Welcome back. Let's go to the next step
in the prompting formula. The limits. Here, the limits
are very important. So you kind of like
restrict chat DVD to get into the outcome
that you really want it. Here you need to
think in three parts. The first one is the format. What is the format that
you want the answer to B? There is no necessity
to repeat it again. If in your action part you
said generate a block, generate at least generate
a script because it already knows the
format that you want. But if you haven't mentioned
it because you make maybe an open
question like, Hey, can you explain me how
the light works? Then? It really, it's
going to give you a very detail and
explain version of it, but it won't be in any format. So if you haven't mentioned it, but it's something that
you want from the outcome, then you should mention it here. Also, you should mention any restrictions here
is mainly about e.g. the length of whatever
response you want. In the case. Let's see, you need a tweet, e.g. you need to say
that it needs to be the exact amount of characters
that you want it here. One thing to keep in mind, according to the CEO of OpenAI, GPT and any API has a very difficult time getting into the amount of characters that you ask it for. It seems like it's some sort of limitation or it's not
actually an imitation, but it's just a difficulty
that the model has. Any like big language model has. So just be aware of this. If you say like 300 characters, it won't be exactly
300 characters. And I've seen this a lot in
my experimentation and in everything that I've done with LGBT everytime I
said like a very, very, very strict
character count. It's not exactly like that. It doesn't meet the restriction, like very, very sharply. So just be aware. If you have a very,
very sharp restriction, it's even more
important that you verify and tried to edit the answer the chat
DBT gives you for the rest of the restrictions you can actually tell it, e.g. don't use certain words or
use certain words or at, or maybe don't talk
about an exact topic or talk about it in a different way that
will follow perfectly. One restriction that
I always love to use mainly for
marketing purposes, is the tone that
I wanted to use. E.g. you can ask you to
be sympathetic or you can ask it to be empathic
or inspirational, or angry or sarcastic or funny, or whatever kind
of tone you would like to give it
to the polls were to anything that is going
to generate for you. And the other one that is very
important to keep in mind, it will be the language
guidance here I included it mainly for
languages that has, again, these strict
differentiation between genders or between
different aspects. It depends to who
you are talking is the way you are going
to talk and et cetera. Or if your brand needs to
be very mindful of these, you need to tell it to chance. E.g. I. Do a lot of words in
German for my clients. And I need to be very specific
in both the gender that I wanted to use and the formality
that I want you to use. So I need to be very specific. Here I have an
example that I did. This is kind of like working on the example I showed
you last time where we asked you to create an script of one short format video and
tell me exactly the script. And what I added here was it, the script should include the
title and the dialogue for each scene with suggestions
for places to film the scene. Okay. So I asked you, don't only give me what I
should do within the video, but tell me where to go to film, where should be the
good location to do it? I also asked you to the video shouldn't be
longer than 60 min. Is this something that
already was there? But I told it as well that
it should be minimum 50 min. I just wanted to know that
this is lengthy enough. Make sure to always
create the dialogue for a female audience
in an informal tone. So you see, I just
put it the format, the restrictions, and
language guidance. And it gave me a really good
example here with the title, even including in
brackets what I should feel out in my case or how I should personalize
these parts. Okay, so we are now here. I'm almost done and completed
our prompt formula, but don't miss out the next
module because I'm going to give you the best team ever, something that you shouldn't miss in all the prompts
that you create. See you there.
12. The Prompt Formula: We're almost done with
a prompt formula, but you cannot miss this part. This is my pro tip and
this is something that has helped me a lot with
everything that I've done, which LGBT has saved
me time and also has given me so much clarity
into my business, into the business of my clients, and into every action that I've taken with the help of GPT. So please, please,
please don't miss it. And let's go ahead. And here is an amazing protein
and this is the last part. This great formula that this
is the cone fear mentioned. The conformation is amazing. I mean, it, it helps you understand how to DBT
has seen everything, but also it helps you not
losing time and just give all the information
that tells DBT needs to create whatever
you need it to create. So basically what you
need to do is use right at the end of your prompt. If you need any more
information to generate the best answer or need
to clarify anything, ask me before
generating a response. Okay. It's amazing
how this works. And what it basically does
is you still eat, okay? You know what you need to do. You note on all my
contexts, you know, who, which role you're playing, you know, everything
you needed to know. Now, if you think or feel that
you are missing something, maybe, you know that you're missing out on all
the tone of my brand. Or maybe, you know, I didn't give you enough
contexts or maybe I didn't I wasn't clear about instruction that
I wanted you to do. Then asked me before generating whatever you think you
should generate, ask me. Because one thing that has
changed if you do that, you need to understand is if it's not clear
about something, is not going to ask you, is going to generate whatever it feels that
it should generate. Like it's not going to tell you a friend that a human being that maybe
will tell you, okay. But I really don't understand
how your brand speak. Can you use let me know what
are the goals of these? Are? What are you pretending to
do with this pose or what is the goal of this article that you
are asking me to create? It won't ask you that. It will basically generate whatever it feels like you
to degenerate and that's it. And maybe it's a crappy. So with this confirmation, we have the entire
formula for the prompt. I really love this formula
and it's been working for me. Wonders. I know it will work
for you wonders as well. So let's use
summarize everything. So first you should ask
to act like something. You should give it a role, a role that guide in what direction you
want the answers to be, then you need to
provide a context which we are calling
it as well training. You need to tell him everything
that it needs to be told. If you don't know exactly
what you should tell it to, to perform an action, you can ask it, okay, I want you to
create a blog post. What do you need from
me to generate that? E.g. then you need to
give it the action. They're very specific and very clear action that
you wanted to do. Generate a blog post,
generate titles. Bounds with me ideas, give me ideas about XYZ. Generate a script, helped me put together a list of tasks
I should do for my business. Helped me understand what are the questions
that I should ask myself if I'm
creating a product, everything that you need to do it and think about
an action verb. What is exactly what you would tell if you are
talking to a human? What is exactly what you
would tell that human to do? Then we have the limitations, the format, the restrictions,
the language guidance. Do we just need to limit how
you want the answer to be? You know, the first thing I told you in one of
the first modules, whilst you need to be very clear on what is the type of
outcome that you want. If you don't have
that clearness, if you don't have that clarity, then you can use
typically to get the clarity by asking questions
and by bouncing ideas. The final thing is
the conformation. This is my, basically
my pro tip to you. The thing that I want you to use every time I need to use, ask until it as if it
doesn't understand anything. Please ask you and do not generate anything if
you don't understand. Okay. So use asked me before
degenerate basically asked me anything before you
do the generation. In that way you are
not going to lose time just getting confused
with other things, other content that
maybe it's not actually fit your needs. This is a formula. You have it right here. And one small thing that I
want you to keep in mind is that chat GBD is
actually that chat. So you need to keep
it conversational. You can use this
formula in two ways. One way you can
craft the prompt, a huge long prompt, as we saw at the beginning, that we can have like an entire
page foolish of a prompt. But it's better to keep
it conversational. As you've seen here in the
examples that I've showed you, I ask for for the role, then I asked a
question than I do. Then I give the training, then I ask the action. Just keep it conversational
so that way it's easier also for me to understand
why I'm giving the chatbox. And maybe if I arrive to an answer that is not
fitting my needs, I can know why exactly was that. Be aware also of
the big prompts. If you are giving
a lot of training, I do recommend you to give a training before and then give the action the limits and all of that because it
can get confused. If you give a huge, huge prompt, it can get
lost in the middle. Alright, so now it's time
for you to take action. I want you to craft
your first prompt. Think about the
lease that we did in one of the modules before. Everything that you
think that will be helpful for you
and your business. And everything that you're
doing the chat UP can help you with and select one of them and craft the prompt with this formula
that I just gave you. If you have any questions, of course, leave it in
the comments below, leave it in the project
asked me and I will give you my feedback in how you
can improve your prompt. But also I encourage you to experiment because this
is the best way you can learn and you can kind of get
the feeling of how Chad GPT works best for you
and the way you talk and the way your business
or your clients work. I gave you the formula. I want you to
experiment with it. So now it's your turn. Go ahead, open to
LGBT experiment, and I'll see you in
the next module.
13. Iteration is Key: Welcome to this module here
we're going to talk about iteration because iteration t, It is a fact that when you
are going to use tragic, sometimes is going
to get confused. Sometimes it's going
to get it wrong. Sometimes it's going to give you an answer that you
didn't expect. And sometimes that's
where it is going to be. The playground. That's okay. That's fine. Because here, one thing that I've
learned, actually a lot, which are the biggies that
you need to keep trying, do need to just keep pushing, keep trying and iterate over
the answer that you got. A lot of times, my
most used prompts is regrade, redo, regenerate. Because I needed to do it again with more information
that I provided with more contexts
or just telling it where it did the mistake. Basically, what I'm
telling you is to use the trial and error. Don't expect chat GPT to
be perfect and know it all and to or to have
it on the get-go, even if you're using
the prompt formulas, sometimes we're going to maybe be unclear on
what we wanted to. Or maybe within
the conversation, we discovered that we
actually want it something else than what we asked on the
first portion of the time, then we have to retell it. Basically, what you need to
do when you're reiterating is tell it clearly where
the mistake or the error. E.g. if you ask for a list of titles and it
didn't give you a list, but it gave you sort of
like a paragraph with, with what it thinks. It could be a nice way
to come up with titles. It happened it happened
before to me too, sometimes it gets
it wrong in a way. You need to tell it, hey, IGBT, I asked you for a
list of titles, not to teach me how
to write a title. Dbt is very clear because
it's very apologetic. So when it makes a mistake,
it's really like, Oh, I'm sorry, I got confused
and made this mistake. I love this way. If you've got Program, I love it because it feels very much like adult
human that it gives, it, kind of feels sorry
for making this mistake. The second thing
also, of course, we need to be specific
on how to fix it. So hey, I didn't ask you for an explanation of how
to create a list. I asked you for the list. So you just you just need
to kind of reiterate what was the format
that they wanted or what is exactly the
information that you want it, or how that mistake
can be solved. And the third thing is that
you shouldn't iterate on the same shot a over two
different topics, e.g. if you ask a great
Instagram post about rubber ducks, right? So then it generates
a block, whatever. Then you ask, oh, okay, then I please generate two
titles to sell microphones. Then it probably
will get confused, but then probably it will
generate the microphone one. And then you ask for
something like, Oh, hey, can you rewrite me the thing about the dogs
and you will get confused. And even I got confused. Kind of like trying to
tell you tell you need to you like keep one topic or one, especially instruction
in one in each chat. So e.g. hey, Chad, DBT. Right now, what we are
going to do is we are going to do SEO optimized my, all my old articles
in my website. So I'm going to pass
you an article and I want this and this
response from you. Are you okay? Yeah. Go ahead and then you
put article and then pass it the article and it
will generate the response. But if you e.g. positive three articles
and then say like, Hey, can you regret it? It will get confused. So you need to be more specific. Don't say just like,
Hey, can you read, write it when you have like 345 different objects
that it did things. If something is
getting very messy, very confusing, some
conversations getting very long. I go to your shoe to open a
new chart, go from there. And that way you will
also sort of like taking your ideas and go from the beginning and
it will be easier and you will have
better results. The last thing I wanted
to tell you here in this lesson is that you
shouldn't get lost. You study buddy to help you, to help your business, to help your clients. But don't forget, your
uniqueness is something that I've been telling you
during the entire class. Don't forget to read
everything that it generates. Don't forget to edit it. Don't forget that you are you, that your job and your touch and everything
that you think and how you believe is very unique and it's something that chat
GPT can not replace. So don't allow the chatbot
to get lost in it. Just go ahead and
read everything and kind of like tweaking
here and there, just become an editor of the work the
tangibility helps you do. And just always, always put your unique touch,
your unique perspective, your unique way of doing
things, of thinking, of things of performing, of creating video,
of creating scripts. Just always, always
keep you inside whatever degenerate with
the help of chat DBT. And the action that I wanted to take it right now
is just take that prompt that we
just graphed it in the lesson before and
iterate over that. If something goes wrong. At some point, if you're
experimenting and you see that sometimes it doesn't
give you the right answer. Iterate, iterate, iterate,
use it, tweak it, tweak the formula and learn
how your language is better understood by chat GBD
and the way you can collaborate with it in
the best way possible.
14. Final Thoughts: Congratulations, you made it. You arrive to the
end of this class. By now you should be up bro, in using chat DBT. And you have right now
the formula to create amazing prompts to help
you with your business, with your clients, with anything that you
want, some help, that you want some assistance, and now you have it for
free with Chad DBT. I'm very, very happy
that you made it this way that you work with
me with an entire class. And I hope you did the project that you
are sharing with us, the prompt that you created, the results that it
gave you an also how you iterate
over the results. What was the answer
that you got? What was that thing that you got help from Turkey
PT, and also, I hope you have now a very good idea on
how to LGBT can help you actually perform your job better and do things
in a better way. This class has been amazing and I wanted
to do it for so long that I am so happy that I am knowing that last, last module. I am sure it's been really helpful for
you to understand it. And I also shared that AI is
something that it arrived and it's going to stay with
us for a very long time, is going to change a lot
in the context of work, in the context of
our professions. So it's very good that
you are taking the action of learning from the
beginning and you're doing all of these work. I know. I am sure if you
are a small business owner, if you're a creative, if you do anything
in the computer, I know child typically
will help you a lot and will make your work
better and easier. And it will allow you to put more attention in
whatever you enjoy the most rather than spending time doing something that
you don't really like. And also, if you
work with clients, it will help you serve
your clients better. If you liked this class, please leave me a
review that is very, very helpful for me, is something that it will
help me a lot and it takes you just a couple
of minutes, tops. You can add escape into GPT
to write a review for you. That's something
that you can do, but I would love if you can
give me your heart on Review. And also, please please
follow me on Skillshare and let me know if you would like any type of
classes around chat, DPT, or anything else, I would be more
than happy to hear you and create classes for you. I will take a look at your
projects when you leave it. So don't forget to
leave it there. And well, I hope to see
you in another class. Thank you. Thank you so much
for being here. I sent you my love from amazing Spain and I
hope to see you again. Don't forget, leave a
review, leave a review, leave a review, leave a
review. Leave a review. Now. Now. Now you have
time. Go ahead. Bye bye.
Alejandra Tello, Digital Content Creator
