Transcripts
1. Introduction: Earning about artificial
intelligence is not a luxury. It's a necessity
in today's world. Regardless of the market, regardless of the industry that you're at or
your career path, you do not want to
be left behind. The world is moving,
the world is changing, and artificial intelligence
is part of that change. In order to learn about
artificial intelligence, there are key essential concepts that you need to acquire. In this current class, we
are going to equip you with essential knowledge to
help you understand the functionalities of
artificial intelligence, how to go about dealing with AI, a GPT, how to structure
proper prompts in order to best out of artificial
intelligence and Chad GPT. And not just that. We're going to provide you
with an exclusive loopprnt to actually help you get up
to speed with the process. You can just simply download it as part of the current class. In addition to, you're
going to be granted exclusive free access to
a prompt generator tool, which will help you reinforce the concept and apply the
learnings to the practice, all of that in this
current class. You do not want to
miss the chance of learning these
essential concepts, essential practices of artificial intelligence
which will surely help you get up to speed up to track and keep up with
all of these changes.
2. Your Project : Your project for the
current course revolves around actually creating
your own prompt, structured in a way
which would get you the best results out of your
interaction with Chad GPT. Afterwards, you
are going to share your prompt structure
with the rest of the community for feedback.
3. What is AI : First of all, let's define what is artificial intelligence. It refers to the simulation
of human intelligence and machines that are programmed to think and learn like humans. What does that
mean? First of all, we take a look at the process of how humans understand
or communicate. First of all, we collect
information, we collect data. Then we are utilizing that
data in the form of training, then we evaluate the data, then we deploy the data and we learn from our
findings and mistakes. The same logic applies for
artificial intelligence. AI works by processing
vast amounts of data identifying
patterns or correlations, which makes the data
useful and predictable. In order to have a
systematic approach to artificial intelligence. First of all, we need to start off by collecting the data. Then inputting the data into the AI model to
train the AI model. Then we're going to
evaluate the output. Then we are going to deploy the solution after collecting
the data and releasing it, and then we are going to refeed the artificial
intelligence model to fine tune the results, and we go from top to bottom. Collect the data,
train the model, evaluate the output,
deploy the results, and then you feed back the entire iteration process
over and over again, till you fine tune and get the perfect result
that you're looking for.
4. AI Vs Humans: Artificial intelligence
refers to the simulation of human intelligence in machines that are programmed
to think like us. First of all, we as humans, we have what we call as natural neurons which
are inside our brain, which talk to each other
in the form of input. Then we have what we
call as a synapse, and it goes to another neuron. So we got the input, and we get the output. So the way we think, we process
information internally. We collect data from
the world around us in the form of an
input, we analyze it, and this translates into
signals which are going to be transferred into an output communicating to another neuron, another neuron, and then you have the cycle
back and forth. So the same logic has been
utilized in order to develop an artificial
intelligence neuron in which the same logic applies. You get an input, then you do
have a series of functions, algorithms, computer
codes which take place one after the other in order
to give you the output. So from a human point of
view, you collect data, you think synapse happens, which is like electrical
signals inside your brain, gives you an output. The same logic applies with an artificial
intelligence, simply put, you collect information,
make it run through a set of code algorithms
to release an output.
5. Models of AI: Models for artificial
intelligence. Mainly there are four of them. You do have what we call
as machine learning, which is using sample data
to run computer programs, to train them, to recognize
patterns based on algorithms. Simply spotting
patterns and data. You got the neural
networks in which you have algorithms
and computer systems which are designed
to imitate the way humans think,
specifically the brain. Then you have what we call as the natural language processing, which is the ability
to understand speech as well as understand
and analyze documents. This is our main
focus, by the way. Finally, you have robotics, which are machines
that can assist people without actual
human involvement. So if you notice all of them, they revolve around
the utilization of computer based machines, algorithms code in order to actually mimic the
behavior of humans. So what are the four models
of artificial intelligence? You get machine learning where we analyze data for patterns, neutral networks in which we mimic the human behavior
and thinking process, natural language processing in which we understand and learn from languages
used by humans and we transfer it into a code
to execute certain tasks. Finally, you have the robotics which are machines
which have been trained to act like humans or
to facilitate the process, which usually involves a lot of human mechanical involvement.
6. AI Through ChatGPT: Now we've learned
some key fundamentals about artificial intelligence. Let's see how they
apply through Chad GPT. Chad GPT, first of all, utilize what we call as the
natural language processing, mainly the NLP model, which is a branch of artificial
intelligence that focuses on the interaction between computers and humans
through language. In other words,
you're not coding, you're not typing
a series of codes. You're actually typing words, language based instructions. Same way you communicate with your fellow peers or colleagues, you input them
into an interface, which is in this
current case, Chad GPT, and based on the NLP model, natural language
processing, it is designed to understand
the context. Are adding it, you're adding instructions, you're
adding information. It will analyze this, run it through a series
of codes in order to generate an output
based on your input, which is within the whole natural language
processing environment.
7. What is a Prompt : Ve seen that Chad GPT
operates on the NLP model, which is natural
language processing, but you cannot just simply
speak to it randomly. You need to follow a
certain structure, and this is what we
call it as a prompt, which is a specific instruction or Q given to a
computer program, particularly within
the NLP model. You are going to input a prompt. It has a certain structure,
as we're going to see, which will be fed into Chad
GPT in the form of an input, then CAT GPT is going to
be releasing an output, which is basically a response based on the input
that you have shared, after which you are going to analyze the response
that you have in the form of a feedback
and iteration process in order to fine tune the input. It's a cycle. You
give it a prompt, which is a set of instructions, which follow a certain structure as we are going to learn. You are going to input
this into CAT GPT. After doing so, Chad
GPT is going to run the series of algorithms and codes within the NLP interface, and then you get an up response. You look at it, if
you don't like it, you're going to find unit, remdify it, fix your prompt, and to input it back again. You keep the feedback
loop going on and on until you get the results
that you'd like to have.
8. Basic Prompt Structure: Get the best out of Chad GPT, we said we need to
understand how to utilize what we call as prompts. Every single prompt is a set of commands or structural
based commands, which follow certain components or include certain components. First of all, they have
an introduction which sets the context of the prompt,
background information. What are you trying
to do to create, to build, to analyze, to learn? Mainly you focus on
action based verbs. Task description. What
are you trying to do? What is the objective?
Create a product, or a list, create a
list, create an agenda. What is the end result? Instructions and guidelines. What would you
like it to follow? Create a list which
is 1010 lines long, or within a certain word limit. Then you give some examples. You could actually include
based on a certain example. You could copy paste that
example and feed it to chat GPTs interface to compare and
draw some inspiration from. And then you could
add some constraints. For example, let's say, create a list for the
design of a website, make sure that it's
one page long, follow the example attached or analyze the following example
and you upload something. Make sure that it should be done within one month's time or
within a period of a week. All of these, these are layers within the instruction
within the prompt.
9. Wrapping Up: So what do you think? I
truly hope that you found it quite helpful,
concise, and joyful. The same way, I found it quite delightful to teach you
in this current class. Make sure that you
follow my profile for the latest
releases and updates, and I'll see you
in the next class.
Engr. Hussein Attié, Entrepreneur I Engineer I Educator
