Everyday AI Podcast – An AI and ChatGPT Podcast - EP 311: Perplexity - How to use it & 5 things to know that will change how you work

Episode Date: July 10, 2024

Win a free year of ChatGPT or other prizes! Find out out.Wanna know a lil secret? Perplexity's a cheat code for growing your biz. Chances are, though, you've got a few things wrong about this AI powerhouse and you're not using it to the fullest. We dish the real 101 on Perplexity and the 5 things you need to know about it. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Ask Jordan questions on PerplexityRelated Episodes: Ep 271: OpenAI Releases GPT-4o: 12 things you need to knowEp 301: Anthropic Claude 3.5 Sonnet – How it compares to ChatGPT’s GPT-4oUpcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTopics Covered in This Episode:1. Overview of Perplexity2. Differences Between Free and Paid Perplexity3. Use Cases of Perplexity4. View and Edit Sources Using Perplexity5. Future Implications of PerplexityTimestamps:01:20 Daily AI news06:30 Perplexity overview08:52 Many hours spent doing work on the open Internet.13:56 Using perplexity to create custom instructions, responses.15:16 Save time by curating and publishing content.19:43 Decent options, limited internet connection, paid plans.23:57 Live explanation and request for audience input.25:02 Analyze company's and 3 competitors' insights30:07 Check sneaker dunk, remove old or irrelevant sources.32:36 Nike Air Jordans: SWOT analysis39:08 Limit search to 5 websites at a time.42:09 GPT-4o update competes with perplexity.44:11 GPT-4o from OpenAI provides valuable customization options.48:07 Website blocks scrapers to protect information.52:16 Fewer clicks on websites, future in AI.Keywords:Jordan Wilson, Perplexity, paid version, free version, advanced models, GPT-4, OpenAI, Claude 3.5 SONET, AI images, SWOT analysis, Nike, Adidas Yeezy, Under Armour Curry, New Balance Kawhi, athletic shoe brands, fertility clinics, SEO, Chat GPT4, Perplexity features, large language models, research tooSend Everyday AI and Jordan a text message. (We can't reply back unless you leave contact info) Start Here ▶️Not sure where to start when it comes to AI? Start with our Start Here Series. You can listen to the first drop -- Episode 691 -- or get free access to our Inner Cricle community and all episodes: StartHereSeries.com Also, here's a link to the entire series on a Spotify playlist. 

This is the Everyday AI Show, the everyday podcast where we simplify AI and bring its power to your fingertips. Listen daily for practical advice to boost your career, business, and everyday life. Meet Firefly AI Assistant, now live and Adobe Firefly, the All In One Creative AI Studio. Just describe what you want to create and the assistant handles the rest, orchestrating multi-step workflows across Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome. The assistant accelerates execution. Do you like wasting time?

Because if you're not using perplexity day to day, you are just throwing away hours a week. It's one of our most used tools here at Everyday AI. And I've loved perplexity since it came out. But a lot of people don't know how it works and don't truly understand what it's capable of. So we're going to be going over that today and telling you how to use it and five things that I think you really need to know. All right. I'm excited for today's show. What's going on, y'all? My name's Jordan Wilson, and this is Everyday AI. Welcome. This is for all

of us. Everyday AI is a daily live stream podcast, free daily newsletter, helping everyday people learn and leverage generative AI to grow their companies and to grow their careers. And hey, such a simple way to do that when, you know, there's a sea of AI tools. And I think perplexity has to be one that you use almost daily, right? So before we get into that, let's first do as we do every single day and start off with the AI news. So the FBI has shut down an AI-driven Russian bot farm on Twitter. So according to reports, the United States Department of Justice, has seized two domains associated with Russians' disinformation campaign that are utilizing a thousand AI generated accounts on social media to spread fake news.

So these Russian actors allegedly linked to Russian state TVRT operated a sophisticated bot farm that used AI to create fake social media profiles as Americans to support Russia's actions in Ukraine. So Deputy Attorney General Lisa Monaco emphasized that the DOJ will not tolerate. the use of AI by Russian government actors to spread disinformation and fuel division among Americans. Also, FBI director Christopher Ray stated that the bot farm aimed to disseminate AI-generated foreign disinformation to influence geopolitical narratives in favor of the Russian government. You know, this is one thing when it comes to Twitter and people say, oh, you know, GROC, you know,

Twitter's large language model, I say, well, you know, there's such a huge chunk of Twitter that is just bot farms from all over the world. So, you know, no big surprise here. But, you know, before the election, I think this is important to talk about. All right. Our next piece of AI news, Microsoft and Apple will reportedly no longer be on Open AI's board due to regulatory scrutiny. All right. So this is huge news, y'all.

So according to reports, Microsoft has decided to relinquish its observer seat on the board of OpenAI just eight months after securing the non-voting position. Also, Apple's plans to join OpenAI's nonprofit board also in a non-voting position have been canceled, as reported first by the Financial Times. So Open AI confirmed the departures in a statement expressing gratitude to Microsoft's confidence in the company's direction. So under the leadership of CFO Sarah Fryer, Open AI is adopting a new approach to engage key stakeholders like Microsoft and Apple, as well as investors such as Thrive Capital. So the new approach includes hosting regular stakeholder meetings to enhance collaboration across safety and security aspects. Obviously, there's a lot of antitrust concerns surrounding Microsoft's deal with OpenAI and those have been growing with UK and EU regulators seeking views on the partnership. And I literally talked about this on the show this week, how this new announcement with Apple getting a board seat was going to open a huge. can of worms with federal investigators. Apparently that didn't take long to materialize.

All right. Our last piece of news. More OpenAI, another big one, but OpenAI is planning to block chat GPT and access reportedly starting today in China. So this newsworthy event revolves around Open AI's decision to block chat GPT access in China and to developers, a country where its services are not officially available, impacting several Chinese startups that are using Open AI's language models through APIs. So OpenAI has initiated steps to block API traffic from regions where their services are not supported with China being one of those big reasons. Users and developers in China have been accessing Open AI services via the company's API, despite the services not being officially available in the country. So this move comes after Open AI's revelation of halting

covert influence operations, including those originated from China, utilizing AI models, for spreading disinformation online. So many are speculating that this actually is hinting that there could be a major release from Open AI coming. I'm not sure if I'm buying that, but who knows, stay tuned and make sure to tune in because we'll be covering that if it does happen. All right, that's enough for AI news.

Let's talk perplexity, right? How to use it and the five things that you need to know. So, hey, super excited to have all of you on here. here live. Yes, if you're listening on the podcast, come join the live stream. It's a ton of fun. Join awesome people like Michael and Brian and Tara and Colby and Cecilia and Jay and Nicole and Woozy and Fred and Rolando. So many people joining us and Christopher. All right. So if you do have questions on perplexity, get them in. I'll try to answer some of them as we go along. But you're going to want to repost this episode too. I'll tell you why at the end.

So here's what we're going to be covering in today's episode. We're going to be going high-level perplexity overview. Talk about the differences between the free and the paid versions because they're pretty big differences. We're going to show you how it works live. So live stream audience, we're going to be tapping on you to do this. Yes, this is unedited, unscripted. This could go anywhere. So get your ideas ready on a very famous product from a very famous company that anyone in the world would know. So get your ideas in now and I'll call for them in a minute. And then we're going to tell you the five things that you need to know about perplexity. All right.

So let's start and go over a perplexity overview. Okay. If you haven't used perplexity before, it is very similar in setup to a standard large language model, right? Even in the user interface and user experience. So essentially, you type in a prompt, right? And hey, for our podcast audience, this might be one of those you might want to click on in the show notes and come watch the video,

but I'll try to do my best to describe what we're showing on the screen. So perplexity is essentially it's set up to look like a large language model, but it doesn't really function like one because it is actually an answer's engine, right? So you have all your kind of saved chats on the left hand side. There is a prompt box like you would see in chat GPT, co-pilot to Gemini, Claude, et cetera. And you can... Get answers, right?

But that right there, that one word is the biggest difference between perplexity and other large language models. Because perplexity is not technically a model. All right, let me repeat that. Perplexity is not technically a large language model. We'll get into the details of that here in a second. But it is actually an answers engine. Okay. So, and this is why I started off the show by saying this.

If you're not using perplexity, you are literally wasting probably many hours. a week, potentially hours a day. And I do think that this is the future of interfacing with information online, more on that later in the show. But essentially, it is an answers engine. So think of something that you may be doing, maybe some competitive research or some market trends, right? How internet browsing has traditionally worked for the past quarter of a century is you put in a query on Google, you end up maybe opening three tabs, you start reading, you get distracted, you click a related article, oh, you actually need to research something else. You go back. All of a sudden, there's 13 tabs open. Okay, you're like, this is fine.

I'm going to go read all these tabs. You know, you go through there, get interrupted, get distracted, go back. Okay, now of a sudden you have 28 tabs. Your browser shuts down because there's too many tabs. You forgot what you started to research to begin with. And you start the process all over, right? And you might spend many, many hours, right, doing this kind of work. Think of all the work you do on the open Internet. And I'm going to talk about what that needs here in a minute. But for many knowledge workers, right, so whether you're reading a long PDF, whether you're reading a long case study, whether you're looking at and in researching, like I said, competitors, market trends. So much of our time as knowledge workers, for many of us,

is spent doing these exact tasks, going through and trying to essentially find answers and to definitively find information online. Traditionally, we've done this. in Google, right? Hey, does anyone want to shout out our live stream audience what your first search engine was? I like asking that. I think I was using like dog pile way

back in the 90s and Ask Jeeves, the original AI agent, right? So perplexity is not technically a large language model. It is a you can connect different models to the internet. And essentially instead of browsing

a handful or doesn't of sites looking for that one piece of information or looking to find trends or whatever it is you're looking for on the open internet. Poplexy literally goes through and looks and does all the research for you. Yes, I use the internet so much less than I did two years ago. I browse way fewer websites and I get way more done. You know, whenever someone asks me, hey, Jordan, what are your favorite AI tools, right? my first is always chat GPT, but usually my second or my third is perplexity.

And I have, I'm not exaggerating y'all because I may or may not be helplessly addicted to software, but I've literally used more than a thousand pieces of AI software over the last like three years, right? I'm always trying them out, you know, reviewing them for our YouTube channel. And I'll tell you this, chat GPT is in its own league, in its own category, right? but perplexity isn't too far behind. And there's a lot of new features being added. Even over the last two months,

we've seen a lot of new features, which I'm going to be going over. And I do think that ultimately, this changes how we use the internet, right? Yeah, going through everyone's, you know, search engines here. So yeah, we got dog pile. We got Yahoo.

We got Jeeves. We got Yash. Who chat rooms. Love it, Tara. Yeah. So now I barely use Google if I'm being honest, right? Like anytime I'm trying to grab information from the web, I always do a split screen. I always have perplexity.

And then I have chat GPT that actually has some great updates that came out two months ago to how browse with Bing works. But I've already covered that. So that is perplexity at a very high level. So like I said, if you do have questions as we go along, please get them in. I'll try to do my best. I'll try to do my best to get to any questions as we go along. All right.

So let's talk about kind of the different main features inside perplexity. And again, right now I'm just showing some slides on my screen. Don't worry. We're going to be doing some things live here in a second. All right. So essentially, let's talk threads. Okay. So threads, think of threads as a chat.

Okay? So if you just click the like how you would in any other large language model, if you click the plus button or new chat, a thread is essentially a chat. So it's a full conversation with perplexity, including your initial questions and your followups. And then that's stored in a library. So you can always go back, right, and find something, a perplexity answer that you did a month ago and continue on in that conversation. Okay. So that is what a thread is. So you can have

a trillion of them, right? You can just do one and keep going on. I would recommend, obviously, if you've taken our prompt engineering course, still works like a large language model. So you should have dedicated threads for depending on the topic that you're researching. Okay. And then there is something called collections. Okay. So collections is essentially a way to organize related threads and it can be shared with collaborators. So yes, just like all other kind of large language models, there's team settings where you can share a lot of these different tools. So collections, also a big thing is you can have kind of a custom prompt for collections. So as an example, I have a collection that I use all the time. So if you read our newsletter and if you haven't,

my gosh, does anyone in our live stream read the newsletter? I'm just curious because I'm a human and I spend a lot of time on it. I write it every day. But one thing that I use perplexity for, an example of collections because you can set essentially a set of custom instructions for your profile, right? But if you use perplexity for a lot of things, so let's say you could tell perplexity, hey, always respond in short bullet points. But then what happens if you want to use perplexity to help you write a long form blog post? That wouldn't work. So a big key element of collections is you can kind of set essentially a specified prompt or custom instructions for anything within that collections.

So in the newsletter every day, we have kind of what's called a try this section. Okay. So what that is is we break down the conversation for the day. You know, normally I'm interviewing a super smart expert. I write it. And then there's three main sections, essentially three main actionable points. And one thing I always try to do is to find research, studies, statistics that go along with those three points. right so i have a collections a collection in perplexity that kind of has a custom prompt and it says hey here's

what you're going to do i'm going to paste in these three main points you're going to go through and research and find me three different examples of recent facts studies figures from 2024 that are supplemental or complementary to these three main points right otherwise i might have to spend every single day 30 to 40 minutes trying to find those. Instead, I can spend about two to three minutes reading through the ones that I think are best, click on them, make sure it's good, and then I can include that, right? So that's just an example of a collection. All right. Then we have pages. So pages are kind of new. So these are essentially in-depth articles or reports that are created from a thread. So you have the option that any thread you

create, you can essentially publish that online and share it publicly, even those, who aren't on your team or who don't even have a perplexity account. So more on that later, but this is actually a huge SEO play. All right. And then last but not least is Cards. So Cards is also a new feature. And if you want to know more about these features, we're not going to be going over all these lives,

live because it could take a while. But we've covered all of these in our YouTube channel, right? Yeah, we have a YouTube channel. It's kind of cringe. I know. I actually didn't pay attention to our YouTube channel until, one of our contacts from Nvidia. So shout out, shout out to her.

Kind of made fun of me. She said, Jordan, you're your YouTube channel. It's kind of not good. So we started paying attention to it. But yeah, we've covered all that on there. But cards are essentially a way to visualize certain data. The problem is, not the problem, but the catch is it can be hard to actually generate those.

And it really is mainly for a couple key data points, such as up-to-date information on financial economic data. etc. So cards are actually a little hard to generate, and it's more random, but I'm sure that will be improving. All right. So now let's talk about the difference between free and paid. And this is huge. And I'm actually curious for our, you know, live stream audience, do you use the free version or the paid version? I think for whatever reason, so many people don't use the paid version, right? Because it's like, oh, here's another $20 a month, right?

My subscriptions are stacking up. I'd say it's worth it. That's just me. I'm not getting paid by perplexity to say this, right? Or open AI to say chat GPT is the best. I care about your time, right? And think of how, right, I kind of describe that common scenario where it's like, oh, you need to research something for work and, you know, all of a sudden you have

a thousand tabs open and you forgot what got you there in the first place, right? Think about your time. How much is your time worth? And I'll tell you, almost every single time that I use perplexity, minimum, I'm saving 30 minutes. A lot of times I'm saving hours. So if you are a business leader out there, think about how much your team's time is worth, right? And also, repost this because we're going to be doing something fun later too, all right? Adobe just introduced an entirely new way to create, bringing the power and precision of its

creative suite into one conversational experience. Meet Firefly AI Assistant, now live in the Adobe Firefly app, the All In One Creative AI Studio. Powered by Adobe's Creative Agent, Firefly AI Assistant lets you start with your vision, just describe what you want, and shape the outcome as it takes form with the Assistant. The Assistant orchestrates multi-step workflows, drawing on 60-plus, pro, Prograde tools across Adobe Creative Cloud apps, including Photoshop, Illustrator, Premiere, Lightroom Express, and more to help bring your ideas to life.

You can also get started with creative skills, a growing library of pre-built workflows for common creative tasks, like batch editing photos, creating mood boards, portrait retouching, and creating social variations. Every step the assistant takes is visible, so you can refine, redirect, or take over at any time. You stay in the driver's seat as the creative director. Adobe Firefly AI assistant now in public beta. See it today at firefly.adobie.com. All right.

Had to grab a drink there. I'm a little thirsty. All right. So the difference is between free and paid. And yeah, I was kind of right. It looks like most people on here our live stream audience are using the free version. All right. So there's some differences.

I have them up on my screen here. But essentially, really what you want is these pro-sergey search. And if you have a free account, you essentially are limited to five pro searches a day. Or sorry, it's actually not per day. I believe it resets every five hours. At least it used to. I'll have to double check on that. On their page, on their pricing screen, it says every day, but I actually don't think

that's what it is unless they recently change it. Okay. But you do have unlimited quick searches. So think of it like this. A quick search is this. It doesn't go through a lot of business. depth. The pro search is much more, much more powerful. The other big difference with the pro, the $20 a month plan, aside from having this much more powerful in-depth search, is you can

choose a model. So essentially, perplexity. So I started off by saying it's not a large language model, right? It needs an engine. Think of perplexity as an answers engine, but it needs an actual engine to run. So if you're on the free plan by default, they essentially have an open source version of Lama that runs it. So it's pretty, pretty decent, right? But if you are on a paid plan, you can use GPT40 from OpenAI. You can use Sonet, Claude 3.5 Sonnet or Claude 3 opus. So you can essentially, right, so think of all these things. A huge, downside of Claude, right, is it's not connected to the internet. But you can, if you have a paid Claude account, you can use your API key and then essentially connect Claude to the internet,

right? So as soon, I kid you not, I didn't always have the paid version of perplexity because it's like, I got enough free, free accounts, right, that I could switch around. But hey, once Claude 35 Sonic came out, I instantly resubscribe. I go back and forth, right? Because I have way too many subscriptions, but I instantly resubscribe to Perplexity Pro because Claude 3-5 Sonnet is really good, if I'm being honest, but it's not connected to the internet. So you can connect it to the internet via perplexity and then use it inside of perplexity. So yeah, you can also do AI images. Me, aside from Mid Journey, I don't really care too much about AI images. So you can only use Dolly and stable diffusion. All right. And then you also get $5 a month worth of API credits.

So that's the difference. Essentially, free version, you don't get a lot of these pro features that are super powerful and that ultimately give you the best and the richest results. And you cannot choose your own model. All right. And you're kind of stuck with the default model, which isn't that great. It's passable. It's serviceable for many things.

But the ability to use GPT40 from Open AI and Claude 35 Sonet from Anthropic is worth the price of admission. All right. Let's get this thing going. Yeah. Woozy said, hey, I'm sure other people can relate here to what Woozy's saying. He said the other day I ended up working with nine different Google profiles open, all on different tabs. I'm guessing you meant to use that many free versions of perplexity, right?

All right. Here we go. So, like I said, the pro search. Yes, it does say it reloads every four hours. Okay, so if you're on the free plan, you get five pro searches every four hours. All right, so get your ideas in, live audience. I need a famous product made by a famous company.

All right, get it in now. I'm going to be sharing my screen. So give me a second here. Got to do some copying and pasting, but remember, we have a worldwide audience. All right. please try to think of a product that everyone knows from a company that everyone knows. All right.

And I'm going to give an example here of a way that we can use perplexity. All right. So in live stream audience or sorry, podcast audience, I'm going to do my best here to try to describe to you what we have going on. All right. So give me a second here, y'all. We're going to bring up perplexity. And we're going to jump in and do this thing live. It's always an adventure when we do these things live, right?

Unedited, unscripted. Could go terribly, but let's see how it goes. All right. So a lot of people said, let's see, Coca-Cola, Microsoft Surface. Let's see. iPhone. Okay, iPhone's a good one. Nike.

Put out a product as well, if y'all don't mind. I kind of like Nike. We might take Nike. All right. And, okay, Nike Ear Jordans. Perfect, perfect. All right. So Kobe, Colby said Nike.

Thanks for the suggestion. And then we had Jonathan with the Air Jordans. Love to see it, y'all. Love to see it. Okay. So here's what we're going to do. Give me a second here. So we're going to say an update to Air Jordans by Nike.

All right. So let's go ahead. and do this. All right. So copy and paste prompts are never the best, but for live, it kind of works. All right.

So let's go ahead and do this. And I'm going to explain to you what's going on as perplexity works. Okay. So I started this off. And again, we're doing this live based on suggestions from the audience, but let me explain to you podcast audience,

essentially what we're doing here. So I'm saying, please take your time and focus on information from 2024 only. Please conduct a competitive SWAT analysis for launching a new product update to Air Jordans by Nike. Technically, it's Air Jordans by Jordan Brand. So we'll actually see if that trips up.

And I know this because I actually, for like 10 years, did a lot of close work with Nike and Jordan Brain. I was out at world headquarters all the time. Shout out to my old Nike and Jordan Brand peeps. So I'm saying before doing the analysis, please first research the company and the product. Identify at least three major competitors who compete with that company and make a similar product.

then provide a detailed SWAT analysis of their strengths, weaknesses, opportunities, and threats, or a competitive SWAT analysis. Include the company SWAT as well as the three competitors. Focused on their product offerings, pricing strategy, distribution channels, market positioning, customer reviews, and recent news or developments related to the product. Summarize the findings for each competitor in a detailed and pinpoint specific manner, highlighting key factors that could impact the product and company's success. Do not include any vague statements or generalized responses.

Every detail should be well researched, analytically sound, precise, data-driven insights, and evidence-based information. However, in your reply, do not waste words and only reply with specifics and details as requested. Please take your time, go step by step, and research thoroughly. All right. So I'm going to go ahead. Hey, a little pro tip here from someone that uses.

is perplexity and large language models way too many hours out of the day. I already, I already have a follow-up, and I don't even need to read what's happening. This is how I recommend everyone use large language models. I'm essentially saying that is not specific enough. Please make one more attempt and re-research this from scratch. Please go to all new sources. Okay. So that's the thing with perplexity, and I'm going to show you why.

But I'm going to say, I'm saying please go to all new sources and redo the SWAT analysis. However, you can still keep. the previous research in mind, but make sure to make a second pass at your research and seek out more detailed, comprehensive, and analytical sources. All right. Essentially, we're just saying, do it again, right off the bat. All right. And I'm going to go ahead and show you kind of live how this works.

Okay. So we're going to look at the first one, and then we're going to look at the second one. So now you'll see right here, this says pro search. And what I like about perplexity is you can go in and kind of check the work, right? So it broke this down into three steps. All right. So it says, step one is research the company, Nike and their Air Jordan product line. Step two, identify at least three major competitors that offer similar products to Air Jordans.

And then number three, conduct a detailed SWAT analysis for Nike's Air Jordan products and the three identified competitors. All right. So again, I do want to point this out there. I cannot show you in a 30 to 45 minute live stream slash podcast, every single feature. All right. So I'm going to go through some of them right now, just so you know, I'm not going to do all these live, but I'm going to show you essentially what perplexity is capable of and why you might want to do this.

All right. So as an example, I'm now toggling down all of these three steps. So I can go in and see exactly what perplexity did at each step. Okay. So for step one, it essentially Googled, right? It Googled all these things, right? So it searched for major competitors to Nike Air Jordans 2024, brand similar to Nike Air Jordans 2024. Competitors to Nike Air Jordans 2024. That's for the first step. And then it read an article from Complex. It looked at a YouTube video for top 10 best upcoming sneaker releases. It went to sneaker dunk, Beast Strategy Hub, Reddit, Complex, etc. So essentially, it shows you,

each step what it searched and the pages that it read. Step two, conduct a detailed SWAT analysis. Same thing. It shows me what it searched. It searched four different terms. And then it went to, it looks like six different, six different web pages. All right. And then, oh, that was the third one.

So, okay, essentially, I can go and look at each different step of the process, see what it searched, and look at the pages. All right. So let's just say as an example. for this one, the second step, I saw that it went to a YouTube video and I can hover over it and I can see the exact video that it looked at. So this was a YouTube video that said, top 10 best upcoming 24 sneaker releases. So let's just say, ah, that could be anything. I don't want that. I don't want that in my, in my kind of reports, right?

So I can go down. Let me see here. I got to go to the bottom. There's a little. lot of stuff here. So give me a second. All right. So at the bottom of the response from perplexity, there's some different options, right? So I can share this. I can rewrite it. So if I click on rewrite, right? That's the thing. If you have the paid plan, I can say, oh, instead use GPT40. My default now is Claude 35 sonnet. All right. So I can share it. I can rewrite it. I can copy all the responses to the clipboard. I can edit the query or the three dots. All right. So the three dots for more options. is where the magic happens. Here's why.

I can click view sources, okay? So I can see all the different sources that were kind of used. So I can go through and say, okay, let's just say there's this, let's look. All right, there's something from sneaker dunk, right? So I can click that and I can say, yeah, that's probably not a reputable source. Or I can click two. And then I can click remove two sources, right? So if you look at the articles or if it looks like for some reason, it's bringing in old

information, that's why I started out that kind of prompt saying focused on 2024 only, right? But if I see it's bringing in irrelevant information or information that could potentially be incorrect, I can go in, click those three dots and remove any sources. And essentially, once I click remove, it's going to restart and rerun it. Okay. So that is the first thing that I wanted to show you. But y'all, yeah, Michelle saying here, you know, Jordan, mind blown as always. Fred, I did not write the prompt that quickly.

I just copy and paste it in. All right. Ah, Monica, I didn't even think Jordan Jordan's, right? This is what I get for not sleeping. I didn't even know. Oh, we're talking about Jordans. My name's Jordan. All right.

So let's go and look at some of the results here. Okay. So again, remember, before the first. first one was already done, I essentially said, eh, do it again, right? I love doing this, huge hack. So, you know, I said, that is not specific enough. Please make one more attempt, et cetera. Right. So now you'll see in the second attempt, it actually broke it down into four steps, right? Because I said, eh, you didn't do a good enough job. Y'all, in any large language model, you should always just have a follow-up, essentially repeating yourself, saying it's not

good enough, telling it to do more additional research, but saying it can use both. I spent thousands of using AI large language models. Huge hack. Just do that. All right. So now I can go down here and let's see the competitive SWAT analysis. All right. So here we have Nike, Air Jordans.

It went through strengths. Let's see, right? Because I actually know a lot. I know a lot more than the average person, right? I was actually lucky enough, oh gosh, to go into Nike's DNA store, which even a lot of employees can't go in there. You can see essentially every single important shoe ever made, some of the most famous shoes ever.

Like I said, not even all the employees can go in there. So I know a decent amount about Jordans and Nike's. Man, I used to just get unreleased shoes mailed to me. That was cool. All right. So it says Nike Air Jordans. So strengths, iconic brand with strong heritage and loyal customer base, collaboration with top athletes and celebrities, innovative designs and technology, such as Nike air cushioning,

extensive distribution network and e-commerce present, strong marketing, and brand storytelling. I'm curious for the weaknesses here. So it says higher price point compared to some competitors, over reliance on basketball and retro styles. That's true. A lot of the other endeavors and other sports haven't been as well received by the market, aside from basketball and retro. Supply chain disruptions affecting inventory and releases, sustainability concerns around

production practices, sure. Opportunities. I'm not going to read these, but you know, you can go and read these or, you know, maybe I'll just share this. I'll create this as a page. So you can all go read this. And then we'll see what it saw as the other four biggest competitors to Jordans, right? So Adidas easy. I'd say that's probably a pretty good competitor, right? Especially as of recently. So it goes through strength, weakness, opportunity threats. Oh, okay. I like this. It went through some key competitive factors. So I noticed in the second one, it only gave me one of the three. Let me see if it did that for the first one as well.

Okay, it didn't. So let me just go ahead and read the other competitors that it found. So it had the Adidas Yeezys. It had the Under Armour Curry. And it had the, let's see, I saw the last one. Where was it? I think it was the new balance. There it is, the new balance, Kauai.

Lenners. So I probably without thinking, right, because a competitor wouldn't be another Nike or Jordan shoe. It would probably be something from Adidas, Reebok, Puma, Under Armour, et cetera. So I don't know if I would have said the New Balance Kauai. I definitely probably agree with the Adidas, Yeezys and the Under Armors Curry. I don't know what I would have picked for the third one. I don't know if it would be the New Balance Kawhi, maybe.

I probably would have said one of the new Puma athletes. that have their own signature shoes, but not bad, right? So I'm not going to go through this, but you can already see. That is one way. And I can do anything with this information now, right? I could say, let's just say as an example, I could say, you know, great, please format this, please format all of this information to be used in a PowerPoint presentation, include all important details.

So I can copy and paste each slide into PowerPoint as an example, right? So that's the thing. And this is something that I think not just perplexity, but all large language models. People just have all this information floating out there, right? So do something with it. Obviously, chat GPT is the best with this because of all the different GPTs, but don't just leave that information out there. Do something with it, you know, ask it to export as a CSB, etc.

All right. I do have one other example. I just wanted to go over this one live. I had this one pre-plan. I'm like, let me think of something kind of random, kind of random, right? So let's go ahead. Let's go ahead. So super random example because I'm like, okay, this is good for researching anything, right?

Perplexia is just great for research. So let's just say that as an example, you're looking for potential leads. So in this example, I'm saying, I sell software to large fertility clinics, right? but please help me compile a list of potential customers. I'm looking, actually my old company, Accelerant Agency, we were actually talking to a lot of fertility clinics to help them with their digital strategy. This is why I kind of pick this example.

So I said, I sell software to large fertility clinics. Please help me compile a list of potential customers. I'm looking for a comprehensive list of large fertility clinics located on the west coast of the United States, right? So let's say that you're selling software for fertility clinics. I know this is a niche example, but I just wanted to give you one. right? And then I'm saying these clinics should have detailed staff information available online, including either their president, CEOs, or lab directors. Please, oh, I forgot. Whoops, I didn't grab

my whole prompts. I got to start this one over. All right. And I'm not going to do the whole thing over from scratch. Don't worry. I'm just going to give you a high level example. So essentially, I'm saying I sell fertility software. I'm looking for clients on the West Coast that I could sell this to. Go look on all their pages. Here's what I want to know. I want to know the clinic name, the location, the city and state, the phone number, the name of someone potentially in charge and a link to their staff or information page. All right.

And then if I scroll down here, look at this. It is already done. It is already done. And it has found me. And let's just go ahead. You know, I'm curious. Let's check this. Because you always have to make sure, are there hallucinations or not?

And that's the great thing about perplexity. And one of the biggest differences, I'd say, and one of the main reasons that people use it especially early on, is because it cites its sources, right? So in that Nike example, when it gave me all that information, I can hover over a lot of things and it gives you these little footnotes. And you can always go and check, right? So now as an example, so it replied back with,

okay, number one, Pacific Northwest fertility, gave me the location, the potential person in charge, the phone number, and then the staff information page, right? So if I jump over here, here's the staff information page. Let's see the person that it gave me, Dr. Laura here. So, hey, Dr. Laura, there you go, Dr. Laura. All right. So pretty simple, right?

And then what I can do, right? Because this is actually a lot. If you were doing this manually, it would probably take you a decent amount of time. Yes, there's, you know, lists that you can go by and do all these things. But maybe you just have a very specific research task like this one, right? And maybe you're like, oh, I want to grow our business just on the West Coast, right? and I have maybe very specific criteria. So maybe if I'm working with an outsourcing company or a list building company,

maybe they don't have that kind of specific information. Well, guess what? Perplexity is doing all this work for me. Look, it's going on dozens of websites at once. So essentially, I just did another prompt. And I'm going to do it again. And then I'm going to show you what we can do with it, right? We're going to let it finish here.

All right, it's done. So essentially what I did two more times, I'm asking it to do five at a time. Any more than that, knowing that, knowing that it's probably going to go to about a dozen websites at once, it might stop working. I think perplexity kind of starts to tinker out once it goes to more than 20 websites or it might not be able to complete your query. So I'm just asking for five at a time.

But then when I'm done, I essentially have a simple prompt that says, great. Now please take your time in researching in five, find five additional clinics. Please do not include any clinics you previously listed. Please do not stop until you finalize a new list of five. Remember, it can be any large clinic that fits my criteria. please take your time, go step by step and research deeply. So essentially, I ran this five times. And I'd say, great, thanks for this.

Now, please convert all of this to a. .c.c. All right. So then it's like, okay, what do I do with this? Do I have to copy and paste it, reformat it, et cetera? No, you can just ask perplexity. It's going to go through now. All of those, it should be 15, 15 for very specific fertility clinics that match what I'm

looking for. I'm trying to sell software to them. It has a contact person, presumably someone that would be in charge. It has a phone number, location, all of this, right? Anything that would be available on their website, I could ask for an email, although it would probably just give me a general email, which could be fine, right? But now it's all done. It's all in a CSV format, right there.

I can copy and paste. Bam, awesome. All right. So that's enough live. I see a couple questions in. Cecilia asking, how long did it take to write that prompt? I wrote it, I don't know, at midnight took me, I don't know, 30 seconds, 45 seconds. All I do is prompt, right?

So my brain thinks and prompts. So it doesn't take me too long. Cecilia is asking here another great question. Thanks for this, Cecilia. Is it possible that given its search abilities, perplexity may find info that is not popular SEO, but interesting. Yes. And you can even, you know, kind of request that, right?

I could even say, you know, don't focus on popular, you know, results that may be at the top of the search engine. So, yeah, you can give it very specific, you know, very specific kind of long tail or long query information that would generally take many hours to research, right? So even to get that list of 15 fertility clinics in a CSV format probably would have, I don't know, taken. an hour, maybe, maybe longer. I don't know, but I just did it live in like, I don't know, three minutes. So y'all, it's all about how much time can you save? All right, Fred, with a question here, other than citing sources, can you summarize the difference between perplexity and chat GPT40?

Great question, Fred. Yes, so perplexity is an answers engine that uses other large language models. It is not technically a large language model itself. So GPT4, like I said, they released an update to browse with Bing that actually really kind of helps it compete with perplexity. But for the most part, you need an engine that makes the car go. Okay? Perplexity is a very specific kind of car that needs a good engine, right?

Whereas chat GPT is kind of the engine in the car or Claude is the engine in the car, right? but the car in perplexity is very unique in that it essentially can browse, you know, 20 sites at once, which is something that most large language models, actually none of them can browse that many. And you can give it specific direction. So it essentially saves all that time Googling. So perplexity is very unique that it essentially can just do all of the targeted research for you.

All right. So I didn't want this to turn into an hour episode. So we're going to go kind of quick here. So we did our overview. We did some live examples. So I hope that was helpful. Now let me quickly give you five important things that I think you need to know about perplexity. Great timing here.

So number one, it's not a large language model. It's actually just an answers engine with citations. Yes. So it does have it's kind of built-in model that's based off of matter. So it's not like a bad model that it has by default. But with the pro plan, you can choose between Claude 3-5 Sonnet. So Sonar Large is, I believe, the default perplexity model that's based off of Meta's Lama 3. Or sorry, not Metas Lama 3, Google's Lama.

Right? Wait, hold up. I don't, I don't sleep. Sorry, that's Metas, Mata. I don't know. I was thinking Lambda. So it's based off of Metas. Lama 3.

And then you have GPT4O from OpenAI, Cloud Free Opus, right? So if you have the paid plan, you can choose what model you essentially want running the car. But the biggest difference is perplexity is an answers engine. It is not technically a large language model by default. It just browses, you know, dozens of websites at one time very quickly. That is what it is for. And when you tap into these other models, if you do have the pro plan, I think that's where the best combination and the best return on your investment and how you can literally,

depending on what you do, potentially save hours within S hours a day. If you have to do a lot of market research, if you have to do a lot of competitive analysis, if you have to do a lot of studying, market forecasting, trends, etc. All right. Did we skip over number two? We might have. Let me see. Give me a second.

All right. Well, apparently it's four things today. All right. Number three, perplexity accepts five. Hey, that's how you know it's done by a human. I thought I had one, two, three, four, five, but I might have deleted number two. So I'll let you know anyways.

All right. So perplexity accepts files. So you can also upload files just like you can inside of chat GPT. Probably chat GPT has the most wide range of file types that it supports. You can upload different files in Claw, Gemini, et cetera, copilot too. So perplexity also accepts files, PDF, text files, CSV, et cetera. But if I'm being honest, it's not. not that great at reading long, like long ones, right? It's just not. It does okay with short ones.

It does pretty good. So, you know, let's say you have a PDF. You can upload a PDF, ask it a question, and it will use your PDF as one of the sources, but then it can still go to all these other sources. Oh, I remember what number two is, y'all. I remember what number two is. So we'll do number two. We'll do it live and we'll show you back the screen here. So number two is you can decide your focus area. Again, I can't cover every single feature in perplexity, but you can set a set of focus for your sources, which is actually really helpful. So by default, it's set to all, right? But you can do academic.

So it's only just going to go through search through published academic papers, writing. So in that case, it's more for generating text without searching through the web, right? So if you're just asking for essentially responses like you would a large language model, there's a math mode to help you solve and find numerical answers. Video mode, which I love. It essentially just searches YouTube videos, right? Which is great, especially for trending things, recent things.

And then socials. So this will search discussions and opinions. So by default, it goes to all, which is why in that Air Jordan example, it went to a lot of web pages, but it also did go to some YouTube videos. So depending on what you're asking, if you do the all mode in focus, it might bring in some research papers, it might bring in some, you know, random internet. articles, it might bring in some YouTube videos, but you can set the focus, right, which is a super, helpful feature.

That was number two. Number three is uploading documents, and it does okay at that. Number four, perplexity is in some hot water. Yeah, they're in some hot water. Just like most large language model makers. And here's why. So Amazon, so I have two kind of screenshots from some, some different articles. So Amazon is reviewing reports that perplexity is scraping online news without

approval. Also, and these are recent stories over the last couple of weeks. Also a fast company article here talking about how perplexity CEO is responding to accusations of plagiarism and infringement. So essentially, right, and this is a bigger discussion, right? But there's been accusations. And, you know, people have kind of shown it to be somewhat true. I'm not going to be the one that legislates that, but, you know, there's been some reports that people are saying, hey, this website is blocking scrapers, right? So a lot of these media companies for a year or more, essentially since chat chbt came out, they block all of these scrapers, right? So if I put something into perplexity, let's say I don't want perplexity and chat chbt and Gemini and Claude

to be scraping my information to be used in their large language models. So essentially, in a file on your website that's called a robot.t. It can say, hey, scrapers, you can't go here, right? Hey, perplexity can't use this, right? You know, but there's been a lot of reports that are saying, oh, perplexity is just using a third party of service that is bypassing all of those rules, right? So generally, these scrapers will follow those directions. So there's been some accusations that perplexity is skirting around some of these basic,

you know, privacy. IP copyright rules and bringing all this information in. All right. All right. So I think there was five. All right. Number four, it has a discover, right? A discover.

I'll show you what that is, which, so it has a Discover tab, which is actually geared, I think, more towards social media sites, which is interesting for a social media site, right? Or for a large language model. So if I click this little Discover tab here, this is all just kind of this trending news. And as an example, you might be wondering why. So here's something that says nothing's CMF big launch day. So nothing is a pretty big, I'd say maybe not big, but a niche technology brand. So this is one of the trending stories.

So it's like, okay, I can click this. And this is, remember when we talked about pages earlier, right? Anything that you create, you can essentially publish as a web page that's open for the web. So this has been a huge play that perplexity has been undergoing. Chat GBTGPT is also doing the same thing. And I'll tell you why. So now just as an example, I'm going to search this, right? Guess what? So now if I search this, this kind of about CMF is big, nothing CMF is, I can't speak, y'all, nothing CMF, their big launch day. Yeah, this is unedited, unscripted. Sorry. But there it is.

the result that perplexity just made in their page, perplexity page, is the top search result. So perplexity is making a huge push to essentially be a search engine because now, guess what? Not only are they taking away traffic from places that would normally be getting traffic and you get all this ad revenue, right? So they're not only doing that with users who are using perplexity, but they are also trying to gain users who are searching for certain things.

to come and land on their page, presumably for ad revenue, right, in the long run and just for user acquisition. So that's a pretty big piece as well. And that leads me with number five, which is related. Perplexity, whether they're saying it or not, is trying to kill traditional search engines and kill traditional web browsing. And I'm here for it. I'm here for it, right?

I do think that this is the future of how to use the web, y'all. I've talked about this on the everyday AI show many times. Using the internet is friggin' unbearable right now, right? Because what's happened over the past two years is chat GPT, their traffic and users have gone through the roof. And as their browse with Bing improves, people are using the internet, much less. Perplexity, same thing. Google, you know, they've been trying to get their AI overviews

formerly search generative experience, right. Same thing. But all these companies are essentially trying to provide you answers. Everyone's trying to become an answers engine, right? Google, Microsoft with their co-pilot, perplexity, chat GPT. So fewer and fewer people are actually clicking on websites. I don't think you should, if I'm being honest. Will a lot of these companies face huge lawsuits? Absolutely. That's another thing for another day. But take advantage of it now. But I think the future of using the web is not doing traditional website searches, right? And for a lot of people, if you're in SEO, I've said this for a long time. Unless your, you know, kind of your main thing is a local service-based business.

I don't know about the future of SEO because I think the future of SEO is people using these answer engines like perplexity, like Microsoft co-pilot, like open AIs, browse with Bing inside of chat, GBT, like Google, whenever they get their AI overview. and SGE and Google Gemini internet connectivity figured out. I think that's the future of using the internet. It saves you time. And when you go on the internet right now, it is terrible because all these major publishers are losing so much money from so many fewer people visiting their websites, they just have to double, triple the amount of ads. I've showed this on the show before.

I've showed screenshots. I've showed live. There's literally websites that go on now if I'm on a smaller laptop. I can't see the content. when I land on the page. There's a banner ad. There's a sidebar ad and there's a main screen ad. There's three ads and I literally cannot get to the content.

I kid you not, right? The internet is becoming unbearable, but maybe that's because it wasn't that great to begin with. So I do think perplexity, if you haven't used it yet, you should, right? Use Browse with Bing the same way. Let me know if you want an episode on that. So that is a very long recap on perplexity. how to use it and five things that you need to know.

I hope you all enjoyed this episode. You all voted for it. We just, hey, we're just here. We work for you. So thank you for those of you tuning in live. Podcast people, I know this one was a little bit all over the place. Hopefully this was still helpful. Again, check your show notes.

You can come back and watch this video. If there's a certain thing, you were like, wait, what did he talk about here? I need to see exactly how this works. I know I don't always do the best job describing things visually, but I still hope this was helpful. go ahead. Anyone who repost this, if you haven't already,

do it. And don't worry, we did the same thing yesterday. We're going to be announcing those winners tomorrow for a free consult. But if you're on LinkedIn right now, you got to do this on LinkedIn,

click the repost. I'm going to send you a video of three additional ways. And one of them I haven't really seen anyone talk about. I'm going to show you three additional ways that we use perplexity, that I think are going to be extremely helpful. So anyone that repost this, it takes you three minutes, y'all, or three seconds.

I was up to like 1230 last night working on this show. Show some support, y'all. Yeah, repost this. I'll send you a video three other ways. I'm not going to post it anywhere else. I hope this was helpful. Please go to your EverydayAI.com, sign up for the free daily newsletter where there will be a much more

summarized version of this long podcast, long live stream. Thank you for tuning in. Hope to see you back tomorrow. And Every Day for More Everyday AI. Thanks, all. Meet Firefly AI Assistant. Now live in Adobe Firefly, the Allman One Creative AI Studio. Just describe what you want to create in your own words and the assistant handles the rest,

orchestrating multi-step workflows across Adobe Creative Cloud apps, including Photoshop, Premiere Express, and more in one conversational interface. You direct the outcome while the assistant accelerates execution. Stand control with the ability to step in and refine at any time. See it today at firefly. adobe.com. And that's a wrap for today's edition of Everyday AI. Thanks for joining us.

If you enjoyed this episode, please subscribe and leave us a rating. It helps keep us going. For a little more AI magic, visit Your EverydayAI.com and sign up to our daily newsletter so you don't get left behind. Go break some barriers and we'll see you next time.