Welcome to the AI Guy Show. My name is Mark Condyles Let's look at Proplexity and see how we
can use this particular tool to really make a big impact in your daily life. I watched a video about the CEO of this
particular company, Proplexity, and it was quite unique in that he really is this
particular tool is a search and also large language model sort of put together. sort of like Gemini, but they are quite a
bit different how they do it versus Google. They believe that they can capture
probably about maybe 5 % of the market share of those individuals out there that
want to really look at optimizing their life. And with this particular tool, I think
they can probably get to that, get that 5 % in the search market itself. So let's dive right into the actual. app itself. I'm gonna bring it up on my screen and
let's just go through some of the how to use this particular tool itself for the
everyday individual out there. So with this, you got the home screen
where it automatically pops up to the actual screen. I actually have the pro version. I liked it so much when starting using it,
I decided to go ahead and spend and pay for the whole year upfront with this
particular application. I found that very useful in many different ways and
we'll go over that today itself. And then you have the discover button,
sort of like what's the top news, what's going on in the world today, sort of like
Google News. And then you have library, so it will keep
up all the threads that you've done. And also you're able to do sort of call
collections, which if you want to do, say this one is doing vacation planning, and
then you can do under that is look at description. And then you can look at prompts that you
are pretending to be a travel agent that will pay for, help provide the knowledge
and information that you need for the trip. So it gives you the ability to sort of
have something separate and you can sort of use this as sort of like a custom GPT
like you see in chat GPT itself. And then all your threads would be there. Then if you scroll down to the actual AI guys show the little wheel. It will give you the ability to use it in
multiple languages. So you have English, French, German,
Spanish, and Korean. And also you have the ability to look at
the different models. So I have the paid version. The other version, the free version gives
you some like five searches with more the detail. But I decided to go with that, the paid
version. And then you have the different AI models
that you can choose from. And we'll look at a couple of these today. The default one, which sort of optimizes
search and also the large language models. And then also you have the experimental,
then the Chat GPT Turbo, you have Claw 3 and then Mistral Large itself. So let's do is... any other things that we need to address
on this particular page. So what we're going to do is jump back to
the thread and we're going to do is keep it pretty simple here. Let's see what the weather is today in
Washington, D .C. So what we're doing is now I put plug that
in and this is the general search for perplexity. And now it's generating the answers. And one thing I like about perplexity, it
sort of gives you right up front. the resources that it's been pulling from. So you can want to look to of those eight
resources, it's of course pulling the weather channel, AccuWeather, Weather
Underground, the local TV station WTOP, and gives you those different options
there of the resources that it's pulling. So with this here, then you can see that
it says weather today, and of course gives it today is March 10th. gives you the temperature, which is going
to be sort of a concise little summary. So with this particular button, which I
really, really like, because many times when I'm doing searches and using these
large language models, in the past, I would have to jump from different large
language model to see what sort of that output would give me. So if you click down here, it says
rewrite, you can then do is then pick this to be rewritten again. So let's do experimental and see if it
even gives a different types of response. It looks like it just gave you a little
bit less, it gave you, this is what's going on with the weather today in
Washington DC, 48 and 35 is low and 7 % chance of rain. So let's rewrite it again with ChatGBT
Turbo and see what the output will be with ChatGBT Turbo. So you can see it's a little bit slower
because it's going out there calling that ChatGPT model through an API. And you can see that it's giving a little
bit more detail in the sense that more about humidity will be around 41 % and 50
% at night and then gave you the sunrise and sunset. The others did not give you that option
itself. And then we'll see what Cloud 3 will give
us. and this response. with their particular model, it sort of
key details out all this information, which for me, I like it this way because
I'm sort of a line by line type of individual. I like to see everything sort of concise,
even though the response output is longer. I do like this particular output because
it sort of goes through and says humidity between 40 and 50, sunrise, sunset, and I
can just quickly scan it. and look at that. And then let's do is look at the last one
and see what it will do for the response. Well, this is very concise and didn't go
until about the humidity, just told chance rain. And so that was all it was giving me for
that capability itself. So let's do is... go and look at image generation. So with this, you have actually three
different types of image generation capabilities. Is that Playground version 2 .5, Dolly 3,
and stable diffusion. So what let's do is go back to this
particular prompt here and do generate image off of weather today in Washington
DC. And we will do is the first one was the
playground 2 .5 and see what kind of response. Well, that is quite interesting for this
particular one. Looks like it's a cloudy ominous over the
US Capitol. And that is probably a realistic depiction
of what it can be. So let's do is change it to Dolly 3. Go back to the thread itself. generate image, photograph, and see what
it comes back with with Dolly 3, and see if it's ominous over the US Capitol with
their particular response itself. It's kind of interesting. It's sort of, even the other pictures,
it's showing snow, which is kind of interesting, but must have been in the
history of Washington DC. Okay, well, this is what is depicting,
Dolly 3 depicted Washington DC and mountains and trees and gave sort of the
synopsis of the weather itself in the picture. So you can see the playground. Okay, so let's do is go back and I think
the last one is stable diffusion and see what it comes back with when you. generate an image on that with the
photograph. So let's see with this one, the different
information that it will provide the user itself. OK, well, I have to say that is a nice one
there, even though I don't think the, which is kind of interesting, the
Washington Monument is not right behind. Looks like the White House, it looks like. or the treasure department. So it is, I don't know where it sort of
put a combination of different landmarks itself. But overall, you can see it's a very
simple tool. One thing it has that I do like, and you
can have it on Android or iPhone apps, iOS apps. You can easily use it. And with typing this, you are able to
attach, you know, whether it's a picture or you're attaching other different
things. you're able to do images, text, and PDFs. It's not able to do like Word documents,
anything at this point. And then last thing I want to finish up
with is the focus. Is that it gives you the capability of
really focusing your search on what you were looking for. So if you're doing extensive research, you
have the ability to say you're doing extensive research in the healthcare field
on a particular diagnosis, you have the ability to go into the academic paper. So if you click that, it will only search the academic papers. And then if you're just trying to write a
short story, you can just hit writing and gives you the capability of doing that
also. And then doing computational knowledge, if
you're needing to do something, some deeper competition type of mathematical
problems, it gives you that capability. And then the ability to go to YouTube. And then one thing I do like is I use
Reddit quite a bit and gives you the capability of searching just a Reddit blog
post. to find information that somebody might
have shared on Reddit itself. But overall, I think what we have seen
with Perplexity is gives you the, I think the biggest variety of the ability to look
at different types of information using multiple large language models, and then
also having the ability from three different image generation types of
software to create images from your prompt itself. Overall, I give Proplexity a thumbs up.
