Transcripts
1. Introduction: If you have been following
the latest trend in artificial intelligence, most probably you've
heard about the introduction of Deep Seek, which is a very powerful
artificial intelligence model. In this current class,
we're going to be teaching you all
about deep Seek. Introducing Deep Seek to you. That way, you have a clear
idea how to use it and the key differences compared to other artificial
intelligence models. Keeping in mind, this is an open source artificial
intelligence model. They can use it anywhere. So it's very important to have a clear idea what how to use it. What are the best cases
for it to be used? And what are the key differences
and key features between the artificial
intelligence models and other artificial
intelligence models based on your own application? We're going to be
covering all of these important
crucial concepts on the latest trend in the
artificial intelligence industry, which is deep seek in
this current class.
2. Your Project: Your project for the class revolves around
creating a comparison, a practical comparison between
Deep Seek and Chat GPT, in which you are
going to be taking the sample prompt
provided to you in the project description or feel free to use
your own prompt, but make sure that
you copy it as is between Deepseek
and Chat GPT, after which you are
going to be sharing the results with the rest
of the community for feedback to help us understand the key
differences between Deepseek and CHAD GPT as
artificial intelligence models.
3. What is DeepSeek?: When it comes to the world
of artificial intelligence, recently there has been a great rise in the
term deep seek, where artificial
intelligence enthusiasts are kind of surprised by the introduction of the new NLP natural
language processing form of artificial intelligence, which is deep seek. So if you haven't
heard about this, most probably you did
hear about deeps. It is the new form of the natural language processing
artificial intelligence. If you're familiar
with chat GPT, it operates under the
same form in which you communicate with it through actual natural human language. You give it a prompt and
actually gives you a response. But what's different is in the architecture,
how it's built. We're going to be diving into it with further more details. But when it comes to chat GPT, the way it was built in terms it collects information is
different compared to deepsk. So Deepsik is considered to
be an advanced NLP model, natural language
processing model, which uses a mixture of
experts architecture. What does the word mixture
of experts architecture, which is the core
difference, actually. The whole rise in the artificial
intelligence industry is because of this architecture. The methodology that deepsk uses in order to
give you results is different compared to HAGPT which is considered to be
the condensed architecture. What is the difference?
We're going to be diving into it at a later stage, but at this current
stage, we need to understand what is deep Seek. So Deep Seek operates
basically under the same concept of natural language processing like HADGPT, but back end, how it's set up, how to collect the information, how to provide you with the results is
completely different. That's why it's been getting
a lot of hype recently because it is designed for efficiency. It's
very efficient. It provides you with
the results with fraction of the resources like other models
have been using. It's faster in response. Basically, instead of utilizing heavy load of resources
to consume a lot of resources to give you the
answer it's actually doing the same action with the fraction of the time
and fraction of the effort. That's why it's been getting
it's getting quite popular at a vast scale massively because it's quite
effective, quite efficient. And to train the
deep seek model, it's actually cheaper
compared to other models. It takes less time to do so. And when you are trying to
get information from it, that's the whole purpose of using artificial
intelligence. You're trying to
actually use it to make your life easier
in various ways, creating articles, writing
emails, and the list goes on. But when you're
using it, the amount of load is different. For example, on Cha GPT, it takes a lot of resources,
a lot of training. Billions and billions
of dollars have been invested into CHA GPT to train the model to deliver it to the current
state that it's at. However, with deep seek, they were able to do so
a fraction of the time, fraction of the resources and getting more effective results. At the same time, the answers
are quite structured. It means you are getting
straight to the point answers. So instead of trying
to re iterate a lot, you're actually getting the
answers straight ahead with less iterations pinpointed with minimal fluff in the process. Why Deep Seek is considered
to be a newcomer to the world of
Artificial Intelligence, where it's competing with previous forms of tools for artificial intelligence
like Chat GPT, because it's cheaper to operate. It gives you a
structured results. It uses a different
architecture, which is the MOE, the
mixture of experts. We're going to be
talking about this to deliver those results. And as we dive into the structure of the
MOE and what it means, you'll get the difference, why it's been getting a lot of hype, why it's very powerful
to use deep Seek. And as we're going to be
saying, this is open source. Consider to hatchP CAPT is
actually privately owned. So you have to actually go on
the website and you have to either use the free
version and then subscribe to get the
premium features. However, DeepsZik
is completely free. You're able to use
it, download it. You're able to have the open
source software program. You can go about tinkering
with the changing that way, it gives you more flexibility. That's why it's been
getting a lot of hype recently. Think
about it this way. You're getting access to a powerful artificial
intelligence tool, which is accessible to everyone, free, takes minimal resources, gives you premium results with fraction of
the time compared to heavy load artificial
intelligence tools, which needs a lot of
training in terms of cost, cost impact. Is the architecture is quite different in terms
of delivering the results. The answer that you
might be getting, which could not be
really accurate compared to the amount of
training invested in it. The other hand, you
do have a model which does not use such resources, and it's able to
deliver great results. So that's why Deep
Seek nowadays has been getting a lot of attention
because of such differences. So that's why if you've
heard Deep seek and it's the whole trend right now where every single AI enthusiast is coming across Deep seek, this is deep Seek. Now, we have a clear idea at this current stage what is Deep seek and why it's
getting such a hype, but we need to dive into
Deep Seek more to understand more intricate details about how it operates and what
makes it different.
4. Key Features of DeepSeek: And welcome back. So we had a brief
introduction about Deep Seek. Now let's take a look at the
key features of Deep Seek, which makes it a great tool, and that's why
it's getting a lot of hype recently, first of all. The whole concept of mixture of experts for
efficient processing. Take a look at this
current diagram. Here. Now, this is your input. This is the prompt that
you put in the interface. You go into Cha
GPT, for example, or Deep Seek and you
type in a prompt. And by the way, we've got some important prompt engineering practices
that we need to be considering as we go about the process of writing
a prompt for deep seek. If you're applying
something in Cha GPT, you could apply
these practices as well to a deep Seek with
a bit of fine tuning, but you still get results. That being said, once
you go about putting the input into the
interface with Chat GPT, it will go through all of the resources it has
to give you an answer. We got a specific
example for this for you to understand
the picture even better. However, with Deep seek, you do have the introduction of a gating mechanism,
a gating mechanism. Think about it this way.
It's like a filter. It's like an intermediate step, which takes a look at your
prompt and then picks up the best resource from
the whole training it took. For example, they called
experts expert one, two, three, and four. It gives them a certain weight. Which one is the better resource to be used to
answer your prompt, and then it gives you the
that's the powerful part. Compared to Cha GPT, ChaGPT will use all of them, whether or not
they are relevant, then to give you the answer. For example, let's say you're working on a piece of writing, and you do want to
translate that piece of writing from English
to ever language. And then you give
this to Chad GPT. It will go through all
of the languages that it knows to go about the
translation process simply. However, with Deep seek, since you've given
a clear instruction from English to a
certain language, it will not go through other sources it has been trained on in order
to give you the result. It will save time, save effort, and it will pick what
we call as the experts. Specific experts,
which are basically the resources to be used in
order to give you the answer. So if you're working on
something related to coding, it's not going to go through the resources related to
languages, for example, right? Because you're
doing coding. So it has the ability to actually pick up the resource related
to your queries. Compared to Cha GPT, Cha GPT will go through all of the models all of
the resources at the same time to give
you the answer based on the model it's using.
That's the whole difference. That's why it's getting a
lot of hype recently with this innovative approach to actually utilizing the prompt input to
give you the result. In addition to the context
length of 28,000 tokens, tokens refers to words,
characters that you input. So when you type a
word, for example, on the prompt interface,
that's a token. Let's say a word,
two tokens, a word. That's the whole
concept of tokens. Anything that you put
on the interface, it could be a hyphen, it could be a dot, it could
be a slash. These are tokens. So hATGPT the window
when you open CHAT GPT, you type the prompt,
for example, the conversation that you're
having with CHAD GPT, it has lower ability to recollect the conversation
that you had with it. However, with Deep Seek, it has the ability to go
through the conversation with up to 128,000 tokens. So that makes it more powerful. Why? Because it's
able to recollect the information that
you have shared with it for a longer period of time. That way, when you're trying
to get answers from it, it has more information within its memory to
give you an answer. So that's why it's getting a lot of hype because
it's actually giving you powerful results with fraction of the resources. It provides you with faster and more accurate AI
generated response, since it has the ability
to handle more tokens, the window when I say tokens, when you're engaging with
the artificial intelligence, whether hat GPT or Deep Seek, you're typing on
a window, right? The screen that you
have in front of you, the words that you're
putting, the communication that you're going through, that's part of the conversation that you have with the
artificial intelligence, whether Chat GPT or Deep Seek. Chat GPT has a certain memory to recollect the information
that you have shared it. Deep Seek has a bigger memory. So that's why it's getting more accurate responses compared to Chat GPT for
certain contexts. It utilizes specialized AI
experts for different tasks. This is the core
powerhouse feature, which has been the
key differentiator compared to other models. When you are trying to get
a piece of information, for example, or you are engaging with a certain artificial
intelligence model. In previous times, you would
go about putting the input. Then the artificial
intelligence model will go through all
the resources it has. It wasn't called experts,
all the resources. Languages, coding,
cooking, whatever it is. And then after going through
all of the resources, it will then pick
up the best output which matches your
desired input, right? Now, with the introduction
of Deep Seek, the process became different where you have the AI experts. These are the
specialized resources. For example, once more, if you are working on a
cooking recipe, right? You input something
related to cook. The gating mechanism we'll take a look at all
of these resources. When I mean resources,
it means the input, the training of the
model, articles, online resources, images,
scripts, whatever it is. I will go through only the resources related to cooking in this current case. Let's say we do have
expert number two and expert number four. These are resources
related to cooking. We're going to use them
to give us the output. So we're not going to be using
number one, for example. That way because number one
is related to accounting. So that way, you are saving
resources to train the model. You're getting more
accurate results because if you're focusing
on cooking, for example, why would you get resources
related to accounting and combine it in the
iterative process? Makes sense, right? So these are the key
features right now which are differentiating
deep seek and one of the reasons why it's been getting a lot
of hype recently. At this current
stage, you're getting a better idea about deep Seek, even though a fundamental level, what are the key differences in terms of the whole concept, the architecture behind Deepseek which is the mixture of experts, which differentiates it compared to other artificial
intelligence models. In addition to the
context length, the conversation window that you have with Deep Seek is longer, which makes it able to
retain information, the conversation for a
longer period of time, getting more inputs from you
as you're engaging with it in order to give you
a fine tuned output. And all of this is
taking place with minimal resources
and free of cost.
5. How DeepSeek Works: So we have learned
about key important features for Deep Seek, but definitely you're curious to know how it really works. So in this current
less, I'm going to be sharing with you some basics about how it works and why it's been getting a lot of hype. First of all, Deep Seek is selectively activating
relevant neural experts. It means we are selecting
specific resources only rather than processing everything
at once like CHAD GPT. Take a look at the schematic. Let's say you're
giving it a prompt. Now, based on the basic
interface, for example, hA GPT, your input will be going through all
of these resources. Consider these experts
to be the resources, books, articles, whatever it is. Now with HAGPT, you
will go through all of the resources
to get your output. However, with Deepsik you do
have a gateway or a router, which gives certain weights to every single resource and classifies them such that it's considered to be an expert. For example, you get cooking, you get accounting, you
got whatever it sciences. So it's like categories, right? So when you do have
a certain input and this input is related to a
certain category or an expert, it will use that resource only. However, in Chat GPT, it will go through
all of the resources, all the training
that it has went through in terms of all
the articles, the books, the data, whatever it's
related to the query or not, and then to find the
best output for you. And obviously, you understand
this current stage. This by itself is heavy. So deep seek reduces the computational load and
enhances the efficiency. So if you're able to use the resource related
to your own query, and get specific results
with fraction of the cost, fraction of the time, avoiding the whole mixing of
random resources. For the company operating the artificial intelligence
model is cheaper, more effective for
you as an end user, you're saving
resources, saving time, saving effort, getting
the information, and the information is more
accurate compared to going through a whole pool of resources and then
to figure this out. Most probably you
have trouble with various artificial
intelligence models that sometimes you don't get the
results that you're asking. You have to go back and
back again and again, iterate the process
over and over again in order to get a certain
form of the result. However, Deep Sea, since
it's based on the MOE, the mixture of
experts architecture, it fine tunes the
answer automatically, finding the results based on the resources and
the training it has, removing everything else to
give you the best output. So how does this
actually look like? Take a look at this
example. So let's consider had GPT first, right? Now, let's say, for example, this is a prom that I'm
what is Hola in English? Hola in English. So this is the dense model, the dense model for hat GPT, means it goes through
all of the resources. It will take a look at
the English resources, Spanish, French, German, Italian, everything that it has been trained on
to answer your prompt. What is pla in English, right? So basically, why do I need to go through French
and German and Italian? I can just simply go to
the English resources and Spanish resources
to answer my query, saving the whole resources, computational load,
time and effort, right? That's why Deepsek
has been coming into the market with
a different approach. It uses a certain sparse
model or the MOE, the mixture of experts model in which OL and English
based on the prompt, what is Ola in English? It means we need to use a certain resource,
English, Spanish. This makes sense
related to the query. So it's not going to go through
French, German, Italian. Well, it might go through
English if it's needed. Let's say, ignoring English
at this current stage. It will use Spanish because
I have Ola in English. I need to translate it based
on the Spanish resources. I'm going to take a look at
the Spanish resources and how are they related to the
English in this current case. So it might include Spanish, I might include English as well. So these are the experts
in this current case, the English resources and the Spanish resources.
These are my experts. The others are not
related to the input, so these experts will
not be considered. Nor is the difference right now. For Chat GPT, it considers all so HAGPT uses all
of the training it has, all of the resources,
whether related or not to answer your
query your prompt. However, Deep Seek has a gating mechanism which actually selects
the resources under the naming of experts based
on your query to give you specific results based on the training and the
resources that has been used. So this is the basic analogy
in terms of how Deepsk works and how it's related to HAGPT in terms of the
mechanism of operation.
6. What can it do?: Back. We have a brief idea
now how deep seek works, why it's getting all the hype. Now let's take a look
at what it can do. What could help us
with as end users. Since the introduction of
Deep Sik is quite recent, obviously there's more room for development in the future, but there are key things that you could use it for right now. First of all, it
could help us with the generation of texts
and assisting in writing, whether you're writing
an article, a poem, a story, whatever it is, it
could help you with this. Debugging and optimizing code. This is considered one of
the core features of deepsk. Its core powers. The writing of code
is very powerful since it's using the mixture
of experts structure. So it focuses only on the
coding applications rather than using other resources
to help you build up a code, making it more
efficient in that case. You're able to write pieces
of code which are short, but the deliver results, and you could debug code
quickly and easily. It helps you solve
mathematical problems with step by step reasoning.
This is very powerful. When you're trying to use
artificial intelligence model, sometimes when you put in a certain mathematical equation, the results are wrong. You're required to reiterate
this over and over again. I've tried this as well,
but I've tried to use different artificial
intelligence models with a different architecture, and if I give a basic equation, just simply gives you out the result without going through the steps and then
you find out that the result if you're able to do it by yourself
manually is wrong. But with deep seek, it goes through the
steps one at a time, and it shows you
what it's doing, and it shows you the
reasoning behind it. And since it uses the mixture
of experts architecture, it has the ability to actually
focus its resources on mathematics and give you the answer, which
is quite optimal. Then we do have
the automation of business processes and
customer interaction. This is very powerful because Deeps has the ability to
be downloaded as an app. At the same time, it's open
source. You can download it. You can use it the
way that you please. So for businesses,
for companies, whatever it is, if you'd like to use Artificial Intelligence, free of cost as of now, they do not have to subscribe
to a certain feature or subscribe to a certain external
party to use the tool. You can just simply
download it and use it. So these are some very
powerful features and the things that Deeps could do which are acting as a game changer in the whole artificial intelligence
industry market. There's the Deep
Seek application. We can just simply download
it from the various stores. And you can use it
as an AI assistant. Even on the application itself, it's named as Deep Seek, your Artificial
Intelligence assistant. You can just simply
download it and on the go, you're able to generate
text, Deb code, solve problems, automate
certain processes, if you download it and you have a certain API to go
about the process. You have the ability
to actually have artificial intelligence
accessible for free of cost and for everyone. That's why it's been getting
a lot of hype recently. These are some of the
basic capabilities that it could do to help
you with on the spot.
7. Why use DeepSeek?: Now, a very important
question arises. Why use deep seek? I'm going to be sharing
with you my insights after going and
using a deep seek. There are key important
features which make it a very powerful
option to have. Well, you do not
have to actually use either artificial intelligence
models, one or the other. You can use them
both, for example, based on various tasks. Now, first of all, it's faster. Since it's using less resources, you're able to come
up with answers on the spot quickly and effectively and efficiently as possible. It's more structured
and accurate outputs. There's less fluff
in the wording, just simply straight to
the point answers which saves you time and gives
you the end result. It improves the efficiency in
handling specialized tasks, specifically mathematical and
application based queries. It's powerful. In that case, you
can depend on it. And these are two
of my favorites. First of all, localization
access is possible. You have the ability to actually download the deep seek
application on your computer localized without
access to the Internet and use the own data that
you have without sharing it through the Cloud service in case you would like to have your own artificial
intelligence on the go without needing to subscribe
or pay a certain fee. You can just simply download
it like a software. Use its capabilities
on your own machine, which is not
available elsewhere. And most importantly,
it's open source. So if you have a coding
background and you're able to tinker with this you have
the ability to modify it. You have the ability
to use the code. It's available
everywhere for deepsk. That way, you have the ability
to use it to share it, to modify it freely, rather than being privately
owned in that case. So these are very important
perks of Deep Seek, which makes it a
strong contender in the artificial
intelligence market. Because think about
it this way, you could get the same results. Get effective results,
structure output. You can depend on it for computational methods for applications and
for mathematics. And then you're able to actually localize it where you can
use it on your phone, on your computer, desktop
PC, whatever it is. And it's open source. You have
the ability to modify it, tweak it, and change
it as you please, and free of cost. So these are basically some
of the important perks for why Deep Seek has been
getting a lot of attention, and it has been ranked as
the number one application on the app store and that
has been downloaded. In the past couple
of weeks or months, which makes it a very
powerful contender, let's say, in the artificial
intelligence domain. And since it has been
newly dispatched, obviously there are potentially more areas for development. Things could be added in the future at this
current stage. Since I've last tried Deep Seek, it has no ability to use voice commands or the
image generation features, which is typified by other artificial intelligence
models like hat GPT, but since it's new, possibly they might be adding
it in the future. But these are very
powerful key features if you're trying to get into
artificial intelligence, trying to learn about artificial intelligence, free of cost, using premium features to actually get you
up running and get important results to
facilitate your day to day activities and workflows, then these are
important perks to consider for why
to use Deep Seek.
8. DeepSeek Vs ChatGPT: Come back. Now let's
have a competition. Let's say let's have
a comparison between Deep Seek and CHADGPT.
Let's compare it. First of all, Deep Seek,
like we have mentioned, uses the architecture
of mixture of experts where selective resources are picked to answer your problem. However, CHADGPT
processes everything on. Every training that
has been went through, every document that has
been provided to it, it will use it to
answer your query, which makes it taxing and
demanding in that case. Deep Seek is more efficient in terms of using
their resources. It can get you the job
done with lower resources, while CHA GPT is more
general purpose. It's not focused on the
application as much as deep seek. It won't get you
results for sure, but for certain
applications like coding and mathematical
applications and solving
mathematical problems, you tend to find Deep Sik
to be more effective. HAD GPT provides
conversational responses, which is a great perk. We're able to communicate
back and forth with CHAD GPT in a more friendly
manner compared to Deep Seek, which is more, let's say, structured bullpoint answers, straight to the point answers, you tend to find
the communication with hATGPT as you're
typing the prompts, the back and forth
communication as part of the NLP model is smoother, but deep Seek is
more structured. So you tend to find Deep Seek zoomed in in terms
of the answer, Cha GPT is more of a general. It goes back and forth in
terms of the iteration until you zoom in and you find the results.
You get the idea. So when you put a
prompt on Deep Seek, it zooms in on the
answer directly. However, with hat GPT, it might require a
couple rounds of back and forth
communication in order to zoom in on the results. So these are some important
features to consider or key differentiators between
deep Seek and CHAD GPT. Up next, we're going
to be learning about when to use each. Like I've mentioned,
you don't have to use one or the other. These are tools for you
that you're able to use. But some of them serve a better purpose for
certain applications, and this is what we're going
to be learning about a next.
9. When to use DeepSeek and ChatGPT: Back. Now let's
consider when to use each of these artificial
intelligence tools, deep Seek and Chad GPT. First of all, let's take
a look at deep seek. When shall we be using it? First of all, you
need a fast response, efficient response
straight to the point with minimal iterations. Deep
See will help you out. You require structured
AI generated outputs. When I say structured, it
means like bullet points, clear steps, deep seek
will help you out. You need optimized coding, technical calculations
or specific expertise, since it uses the whole
architecture of MOE, which is mixture of experts then if you have certain
technical queries, something let's say
requires calculations, deepsk will help you
out because it's able to focus on certain resources to help you with a query
compared to HAGPT which focuses on all of
its training resources. Then if we take a look at the
other end of the spectrum, we got Cha GPT, and when
shall we be using it? For example, you need
general knowledge and conversational AI. You're writing an article, you need some sort of tips. You're trying to modify
something, you need a plan. You need a strategy,
marketing campaign, strategy, business plan. Back and forth communication with artificial
intelligence like Chad GPT, since it has a pool
of information, it will get you results based on a conversational methodology rather than just
simply structured one time shot kind of result, you're able to pick
up and tinker with its own training to extract information
compared to Deep seek, which is more of a focused ends. Then you require creative
writing and brainstorming. This is very powerful
for Chat GPT, since Deep Seek is
more zoomed in. You give it a prompt query, finds the resource related to it and investigates it further
to give you the result. However, if you're trying to
brainstorm creative writing, coming up with an idea, a theme, a project, for example, AGPT outperforms deep Seek because it has access to all
of the resources at once, so the whole concept
of creativity is stronger compared to deep Seek. So these are key
areas that you need to consider when you're
trying to use an AI tool. For example, you have a mathematical equation that you would like to solve Deep Seek. You're planning on
debugging a code, Deep Sik. You would like to write an
article 0R blog post HATGPT. You'd like to come up with a marketing campaign
plan, hATGPT. You would like to come up with various inspirational ideas, quotes, wherever it is, something related to
creativity, hATGPT. You would like to use
something related to debugging
application, Deepsk. So the criteria to
keep in mind is, if it's something
specific that I need to be working on,
I'll be using Deeps. If it's more conversational, I need to get ideas, find informations,
iterate back and forth. To get to a certain
result through accessing a pool of
knowledge, HAGPT. Deep Seek has access to a pool of knowledge,
but the mechanism, the architecture
selects certain pieces of knowledge and resources named as experts to
answer your query. So as you have understood so far the background architecture
that makes them different, you have a clear idea
which one serves your application better such that in order to get
certain results, you have a clear idea whether
to use Deepsek or CHA GPT.
10. Limitations of DeepSeek: Now at this current stage, Deep Seek is still in the development process,
it's still growing. So there are certain limitations
we need to keep in mind. First of all, it may
not be as versatile and creative in whiting
compared to Cha GPT. Let's say you're accessing a pool of knowledge
through Ti GPT, Deepsk does not
have that training yet to give you more creativity in the whiting part or the whole creative
approach kind of thing, whether coming up with videos, trying to brainstorm
ideas, creating some plan, something related to the
whole access of resources to come up with a plan rather
than specific answers. So it might lack at
this current stage. It's still evolving since
basically they have been smaller integrations
in the public. Less people are using it so far, compared to Chachi PT, which has been around
for a couple of years. So it's still in the
training phase, right? It's still in the
development phase. It's doing impressive results, impressive outputs, but
there's still room for growth. So at this current stage,
if you are using it, it's still in the
development process. That requires better
understanding to how to go about the
mixture of experts. This is very important as
part of prompt engineering. When you have been using different artificial
intelligence models such as HAT JPT, you've been using a certain
approach to tinker with the artificial
intelligence model to get your answers as part of prompt engineering
practices, right? But since the architecture
is different for deep seek as part of the mixture of
experts, architecture, the approach to get the results
that you want needs to be slightly modified
in order to make sure that you are actually
able to get the answers. Based on the best
potential output. Since the architecture
is different, the way you communicate with the artificial
intelligence algorithm and model should be different as well,
right? Makes sense. So since the whole
technology is quite new, there's some room for
tinkering to find out what are the best prompt engineering practices in order to make the best of such an artificial
intelligence model. But at this current stage, if you go about deepsk
and by the way, this is simply a screenshot
of the interface for Deep Seek looks quite
similar to CHAD GPT, where you do have the message
or the prompt window, then you do have various
options or models to go about the prompting is
something to be covered. But it's very
important to keep in mind that as you go
about deep Seek, it's still in the growth phase. So you might find that
every single day, there are certain
additional perks, add ons, certain features, certain models that have
been integrated into Deep seek to include certain features which
might be in demand. So at this current
stage, there are certain limitations because
still in the growth phase, but who knows in the future, it might get more developed
with more features, with more variations
to the tools that you could use within
deeps compared to Chang EPT. But from your end, this requires some
training as you go about the prompt
engineering part, if you're accustomed
to a certain way of communicating with different artificial intelligence models, you need to test with Deep Seek how to go about
communicating with Deep Seek in order to get better
results because the same approach that
you have been using might not work as is with this
model ficial intelligence.
11. Wrapping Up: So what do you think?
I truly hope that you found the class helpful
if it helped you get up or running
with a deep seek as a new introduction into the artificial
intelligence industry. It's a job well done. Make sure that you
follow my profile for the latest
releases and updates, and I look forward to receiving your feedback on
the current class, and I'll see you
in the next class.
Engr. Hussein Attié, CEO I Engineer I Educator
