Everyday AI Podcast – An AI and ChatGPT Podcast - EP 243: 5 Simple Ways To Use Generative AI Every Day

Episode Date: April 4, 2024

Win a free year of ChatGPT or other prizes! Find out out.If I had to count, I probably use Generative AI at least 100 times a day.  Some are advanced. Some are super simple. I'm going to be sharing some of the simplest ways that I use Generative AI to create Everyday AI and tell you the ways that you can use my same tactics to start winning back time today.  Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageLearn more in today's newsletterJoin the discussion: Ask Jordan questions on AIRelated Episodes:Ep 189: The One Biggest ROI of GenAIEp 197: 5 Simple Steps to Start Using GenAI at Your Business TodayUpcoming Episodes: Check out the upcoming Everyday AI Livestream lineupWebsite: YourEverydayAI.comEmail The Show: [e-mail] with Jordan on LinkedInTimestamps:01:30 Daily AI news07:17 Knowledge workers adapt to succeed in AI era.10:44 Behind the scenes of Everyday AI production.18:21 Steal this method to learn about your career.19:10 Customize Google search for efficient news updates.23:01 Using shortcuts to gather information efficiently.27:15 Using CastMagic and its features.29:13 Automated transcription33:57 Microsoft 365 Copilot helps with team work.37:05 Addicted to Chrome, but now liking Edge.40:20 Request to find specific generative AI statistics in PDF.44:17 Chat prompt reused for specific AI training.45:02 Optimize chat model use with refined method.50:12 Use perplexity for efficient research queries, not Google.52:03 Generative AI can save knowledge workers time.Topics Covered in This Episode:1. Practical uses of Generative AI2. Tips for saving time using AI3. Summarizing web pages for quick updates4. Organizing and training ChatGPT for specific tasksKeywords:Microsoft 365 Copilot, Microsoft 365 Teams, Cast Magic, Chat GPT, Microsoft Edge, Chromium, Chrome Extensions, Copilot Integration, Perplexity, Google Search, Generative AI, McKinsey Digital, AI News, Amazon Web Services, Voila, Organizing CSend 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 in 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. I use generative AI at least 100 times a day.

I mean, it's 7.30 a.m. my local time, and I've already used it at least 50 times today. And some ways that I use generative AI are very complex. And some ways are extremely easy. And today I'm going to share five simple ways that I use generative AI. every single day and how you should be using them to. All right. I'm excited for this episode. What's going on, y'all?

My name is Jordan Wilson. I'm the host of Everyday AI. And Everyday AI is for you. It's your daily live stream, podcast, and free daily newsletter. Helping everyday people like you and me, not just learn generative AI, but how we can all actually use it. Yeah. And we're going to be talking about five simple ways to actually use it today.

But before we do, as we do every day, let's start off with. the AI news. All right. And you know what? For all of you joining live, you know, thank you, Rolando and Tara and Brian and Jay and everyone else. I have a little poll for the AI news today. So as I read it, let me know if what you guys think.

Should I not read the AI news every day? Should I make the segment shorter? Is it the right length or longer? Just go ahead and drop that in as we go over the AI news for today. So New York City's new AI power chat. bot is being criticized for dispensing this advice. All right. So the AI powered chatbot created by the New York City to assist small business owners

is facing criticism for providing inaccurate and harmful advice, including misstating local policies and suggesting actions that could lead to legal violation. So despite acknowledging the chatbot errors, New York City Mayor Eric Adams has chosen to keep the tool on the official government website. emphasizing the need to address technological challenges. Experts warn about the risk of deploying AI systems without proper oversight, highlighting the importance of implementing safeguards to prevent incorrect information. All right.

Our next piece of AI news, we got two things from Google here. So number one, Google has introduced AI-powered reply suggestions in Gmail for Android users. So Google is reportedly developing a feature. for Gemini advanced subscribers that will leverage the Gemini Ultra AI model to suggest replies for emails in the Gmail app for Android. So if you're an Android user, you might not have to be spending time reading and replying. You might just be able to hit a button soon. So the AI tool is expected to save users time by providing custom reply suggestions based

on the email content visible in a panel labeled reply suggestions from Gemini within the app. So while the applies may not fully utilize the generative AI capabilities of Gemini, they do or will offer quick and convenient responses directly above the text field, enhancing productivity. That's what Gen AI is all about, right? All right, last but not least, this one's a big one, y'all. This one could shake up SEO and how we use the internet. Nothing new here, but some of the first official reporting.

So Google is considered charging for premium AI features in search engine, according to a Financial Times report. So Google is reportedly contemplating offering premium AI-powered search features through paid subscriptions a significant shift from its ad-funded model. All right. So these new features could include AI-generated answers and snapshots, which are more resource-intensive to provide than traditional search results. This move comes in response to the competitive threat posed by both open AIs, chat GPT,

and perplexity, which we talked about. out in our AI in 5 on Monday, and it offers comprehensive answers without the need to click through on the links. Yeah, that kind of goes against their main business model. So as an example, Microsoft, through its partnership with OpenAI, has already integrated GPT-powered search and chatbot features in the Bing, co-pilot, but this has not significantly impacted Google's market dominance. So concerns have been raised about the potential impact on Google's ad business.

If AI-generated answers, reduce the need for users to visit advertisers' website. Yeah, we've talked about this many times here on the everyday AI show and, you know, talking about how these big companies are eventually going to have to shift their models, right? So much of, you know, search engines, clicks or their revenue come from ads. So as you see companies like perplexity, you know, which we're actually going to be talking about today, make that process so much easier. They're going to have to figure out other ways to monetize. And it could just be paid subscriptions. All right. Hey, a couple people voted on the poll.

So, you know, even if you're listening here on the podcast, let me know, do you like the AI news every day? A lot of people do. But let's talk about actually using generative AI. All right. And thank you all for the feedback. This is something that we hear from people who read our newsletter, from our live stream listeners, from our podcast audience. Something we hear all the time is, hey, Jordan, you know, learning a lot. This is great.

but, you know, I want more ways, more practical ways, more easy ways that I can start using generative AI in my day-to-day life. And like I started off the top of the show saying, it's an understatement to say that I use generative AI at least 100 times a day. I would venture to say most days it is hundreds because I have usually on my screen multiple generative AI programs open at all times. I have Gen AI powered Chrome extensions. I'm using it all day. All right. So today I'm going to share five very, very simple ways that you can use these tactics too.

And I do want to point this out. These are simple ways, right? I use generative AI in very complex ways every single day, you know, creating pseudo agents within large language models, leveraging automation, third-party APIs, very complex, right? So I just want to put that out there, right? Like people are going to be like, oh, Jordan, this is the simplest thing ever. That's the point. That's the point.

So whether you are, you know, a Gen AI expert or you are brand new and have never used it, I hope today's show gives you value. But the thing that I really wanted to start with is if you're a knowledge worker, you need to rethink how you work. All right. So a knowledge worker is probably most of us, those who sit in front of the computer all day or most of the day, right? We are reading things on the internet.

We are researching. We are creating on the internet, you know, documents, presentations, sales call, analyzing data, designing, right? So if you spent the majority of your workday in front of a computer, you are a knowledge worker. You are being paid for your expertise in being able to take knowledge that, you know, you have read, you have learned, you have created previously, and applying it to new opportunities. That is what a knowledge worker is. That is what so many of us here in the U.S. and so many people in our audience are doing, right?

So you need to rethink how you work, which I know sounds weird. But if you really want to survive and thrive in the age of AI, you need to drop good habits. Right. We always think about, you know, growing our company, growing our career. is you need to drop bad habits. But this is the opposite. And I know this is going to sound weird, but I want us first,

before we get into these simple, you know, tips to use generative AI. You really have to change the way you think. You have to drop good habits that have maybe led to your company or career growth. Because what we do is a knowledge worker, maybe the reason that you've been promoted multiple times over the decades

may not matter anymore. Right. Your expertise, your ability to recall knowledge from, you know, a very successful career may not matter in the same way anymore or it may not matter as much because of our use of large language models. Right. So one other thing here that I need to start bringing up on this show more. There's a very famous McKinsey Digital Report talking about the impact of generative AI. And I just want to talk about one specific thing in the report. All right. And I'm going to read this out loud and we'll link to it in our newsletter.

So if you're not already signed up for our newsletter, make sure to go to your everyday AI.com. So this part says current generative AI and other technologies have the potential to automate work activities that absorb 60 to 70% of employees time today. And then they're talking about in contrast, we previously estimated that technology has the potential to automate half of the time. All right. So previously McKinsey had a study.

that said 50% now they say 60 to 70% of our time can be automated through generative AI. And I'm telling you right now, this study is very wrong, right? Obviously, it's from June. So I'm not, you know, harping too hard here on McKinsey. But I would venture to say easily 90%. Not exaggerating, y'all. I've been very lucky enough to talk to some of the smartest minds in AI across the globe from the biggest companies in the world, you know, Nvidia, Microsoft, IBM, AWS, right? I'm lucky enough to get to talk to these people. I can see where we are heading.

Very easy to say that 90% of the manual task, this manual knowledge work that we do in front of the computer is going to be automated via generative AI. So you need to learn how to implement and change the way that you work. Enough wind up. Let's talk about the five simple ways. All right. You guys ready? Hey, for our live stream audience,

let me know. Let me know. You know, you know, Tara's said, this is the new norming. I like that. Let me know. Are you guys ready? I'm ready for this one.

And like, you all have asked previously, like, hey, let me know behind the scenes of how you do this every day, right? People ask, how do you do a daily podcast newsletter, live stream, you know, how do you do this every single day? Well, here's some of the behind the scenes, some of the ways,

and this is just the, the simplest ways that we use generative AI here at everyday AI. All right. And I'm going to try to do a lot of these things here live. All right. So please, please bear with me a little bit because I'm going to be sharing my screen. Sometimes this goes well. Sometimes it doesn't.

So let's see. So number one, all right, reading through AI news headlines. This is the way we use it with chat GPT. All right. So let's go ahead. I'm going to give you all, hopefully this works here. So I'm going to give you all a live look at what I do every morning. All right. So I woke up, I might have hit snooze on my alarm once or twice. I woke up at about 510, 510 this morning. That's why I said I've already been using generative AI for half the day. So let's go ahead and take a look. And this is how I, you know, one of the ways that I use chat GPT. I use chat GPT literally every single day.

It is by far the best and most powerful generative AI tool, I would say. All right. So here's what we're doing. We are, I have a chat GPT plus account, right, $20 a month, which allows you to access GPT for turbo as well as GPTs. All right. So I haven't trained this chat. All right.

So if you've taken our free prime prompt polish course, you might be like, Jordan, why? Well, I want to show you this from scratch. So we're not going to go through the whole process. I'm showing you a very, very simple prompt. But the first thing I'm going to do is I'm going to mention a GPT. Okay, so what that means is now I'm talking to web reader a GPT. The reason being, and this is so important when you're using a large language model, and I'm going to spend a little more time on number one because I think it's important.

All right. Yes, chat GBT is technically connected to the Internet. Okay. And why do you need to connect a large language model to the Internet to the Internet? And this is why I rarely use Anthropics Claw or even Gemini, because Gemini from Google actually has internet problems. You need to be able to get the most up-to-date and accurate information when you're using a large language model. So chat GPT, as an example, has a knowledge cutoff of April 2023. So what that means is all of the information, your best case scenario when using a large language model,

is you're going to get information that is more than one year old. That's bad. It's dangerous. That's how you get hallucinations. That's how you get inaccurate information. So I would say 90% of the time when I'm using chat GPT, I am connecting to a third party. We used to use plugins, but those have been depreciated.

So I'm using a third party application to connect directly to a specific URL. All right. So I'm going to go ahead. I'm going to click this prompt. We're going to let it go live. All right. hopefully let me know, let me know live stream audience. Let me know if you can read this.

I'm going to try to zoom in a little bit. Hopefully this works fine. All right. So I'm going to go ahead and show you kind of exactly what I'm doing here. I'm not going to read this whole thing because it's kind of long, but I'm zoomed in a little bit. So I'm essentially saying, please summarize this webpage in a witty voice. All right. Let me show you the web page and why I'm specifically using this.

Okay. I have to read so much AI news. every single day to keep up in conversations, to know what matters for you all, right? So this is the URL that I shared. All this is is it is a Google news search for AI. And you'll see here, if I scroll through, there's probably about 50, probably about 50 news articles, right? So what I'm doing in this prompt is this is the URL.

I share at the very bottom. So essentially, I'm asking the GPT. web reader to visit this exact URL that contains, and this link updates itself every day, according to the most timely and relevant AI news. Okay. So then, you know, I'm not going to read this whole thing, but I'm essentially telling it who our audiences, right, business leaders in the U.S. who are trying to learn and leverage generative AI.

And I'm telling them what our audience cares about, okay? I give it an example of exactly how it should be summarizing all of these different, you know, 50. And I'm just asking it for 12 different recaps. Okay. So normally most days, I'll do this with a Google AI News search link. I'll do it with a Bing AI News search link. So I get a nice variety.

And I will read, you know, at least 30 of these quick little news summaries every single day. All right? Because for a couple of reasons, I host a live stream. y'all every single day for almost a year now it's not easy right and when people have questions I like to have answers I I'd like to keep up I like to stay ahead all right so now if we're we're looking at our results here and if you are listening on the podcast this is one of those you might want to check the show notes and come back and watch the live stream so you can see it but I'm going to do my best to describe it okay so now it's the results here

Let's see if it actually does 12. Probably not. Sometimes it caps out. All right, it actually gave me 12 recaps. Each recap is about four to five sentences long. So it didn't get it exactly right. That's okay. This is actually a somewhat complex prompt.

So now what I can do to start my day, I do this, I get my coffee. I'm waking up, right? And I'm reading it. So this is why I do it this way and why you should too. I'm giving it information, right? And so it is doing so much of the work that I would have to be doing manually, okay? Because it is putting all of this into perspective for our audience or for me, as well as I gave it instructions on exactly how to write this, right? I said, you know, act like you're a TMZ reporter, you know, write it in a witty and engaging way that gives context to our readers, who are U.S. business leaders.

right and talking about specifically how it can grow their companies and grow their careers. So for each and every one of these 12 recaps, it's already done a lot of thinking for me. It's already pulled out the important information and spit back what I actually care about. All right. So that is number one. All right. So now let's jump back in. I have a screenshot here just in case it didn't work, but luckily it worked.

All right. So now let's talk about how you should do this too. And Jay, Jay, thanks for joining this live. I don't know if chat GPT watches TMZ, right? Maybe. But obviously, you know, TMZ is a very popular show. It's in, you know, the 1.8 trillion parameter large language model that is GPT4 turbo. So that should be, that should be, you know, reflected in the response. So here's how you should do this. Obviously, you don't need to do this for AI news. But what you should do is you should steal this method to learn daily about your career or company path. All right. So as an example, as an example, let's say that you are in real estate. I'm pretty sure, you know, Tara here who joins the show almost every day. I'm pretty sure Tara's in real estate, right?

So as an example, maybe you spend 15, 20, 30 minutes a day. Actually, I'd love to hear how many minutes a day or how many hours a week do you all spend to keep up with industry news, to keep up with industry insights to stay abreast to what's going on? Like I said, I probably spend way more time than the average person. But let's say you're in real estate. And you always want to know what's going on. What's the latest?

All right. So what you should do is, is, you know, whatever that query is in Google, real estate news, or maybe you're in FinTech or, you know, whatever it is. You tweak that Google News search or whatever it is. Maybe you want to do it once a week so you can change the, you can change the timeframe to the last seven days and that'll change the URL. And then you put that in and you put it into chat, GBT, and you do just like I did. You give it context. here's who I am. Here's what I care about when I read the news, right? Here's an example.

It should be written in this format, right? And right away, right away, you are instantly saving so much time. Not only you're saving so much time, that's brainpower. Right? You don't have to read all these stories and think. How are these tied to my company, my career, my profession? You tell ChatGBTGPT that up front. You give it an example of how a news story that appears in these results should be written or should be reframed for you.

Y'all, when we think about changing how knowledge workers work, this is top of the line. Yes, it's low-hanging fruit. Yes, this isn't, you know, we're not tapping into automation and APIs and, you know, JSON and like doing all these advanced things. This is simple. But the thing with generative AI is winning back your time and also being more effective with your current time. All right. That's an important thing. Let's go on to number two.

All right. Summarizing web pages into scripts. All right. So this is exactly what I do every single day. All right. So full disclosure, I always read. I always read every single article that I end up. talking about on our news, right? Because I want to make sure that I truly understand it.

You know, in most of these articles, and I'll tell you this, right? So I'm sharing this now on my screen. Give me a second here. There we go. So, you know, here is the article I read this morning from Reuters about Google plans to charge for AI powered search engine. Okay. So I read this. But if you see my screen, this is pretty terrible, right? Because of AI search, things like Google's SGE search generative experience, things like perplexity. Now chat GPT from OpenAI is creating more sources. So they face less lawsuits, right? But pages, web pages are getting harder and harder to read.

Look at all of these ads. My gosh. So like so many ads. It's so hard to read these pages. Anyways. So I go through and read the pages, all right, to make sure I understand. And it's hard sometimes. So then I use a Chrome extension here called voila.

So I'm not sure. and please let me know live stream audience. Hopefully you can see this. I never know if you can see Chrome extensions when I'm sharing my screen. Hopefully I can't. So this is a Chrome extension called voila. Okay. So what this does is it's powered by GPT.

There's a lot of other features. I probably covered this once or twice before. All right. Also, you can choose your model. You can choose to have it also connect to the internet. Okay. The thing that I like about this is, You can have pre-built prompts.

So similarly to what I did inside of chat, GPT, you can save a prompt and have it reflect your current page. All right. So here's what I'm doing. So I just clicked one of my pre-saved prompts. All right. So now, sometimes I will write these bullet points myself, right? Sometimes, if I'm being honest, if I'm running out of time like I was this morning, because I put a ton of work into today's show, I might copy and paste this. And as I'm reading the news, I might reference this. Sometimes I write it myself just with quick bullet points. Sometimes I'll do half and half. Sometimes I'll use this, you know, 80%. All right. But now what just happened? And hopefully you can see it on my screen here. It's kind of broken this down, this long article, that I read, but man, by the time I read this with how many ads there are, there's pop-ups,

I'm axing out of them, it's hard to pull this information, right? So there we go. That is number two. Thanks for letting me know, Jay, that yes, you can see it. Monica, I agree. This is incredibly helpful. All right. So that is our number two.

All right. let's jump back. Oh, well, here's how you should use it. All right. So you should save time and eye strain by training a web page summarizer based on your context and needs. Okay. You can use the Voila, you know, app. There's plenty. If I'm being honest, there's probably at least 50 Chrome extensions that do this exact same thing. I use voila. I like it. I've tried a lot of different ones for my, for my purposes, this works. What I really like is I can share these with my team. I can tweak them.

I can create 50 different ones. I can have a library of prompts depending on what I'm working on. But you should be doing this too. All right. So think. And let me also put this out there. Almost everything, almost all of these five things today, I could do in Chad Chibati, right?

But I'm wanting to show you different techniques because you're probably thinking, wait, why don't you just do this in chat GPT with the web reader? Well, I could, but so much of the time, it's all about speed, at least for me. I work with speed and efficiency, right? So for me, sometimes if I'm on a page, it is just faster for me because I can get that in two clicks without ever leaving. Okay. If normally I do have a chat GPT up on my screen, but if I'm on a laptop specifically, right, I probably don't have that chat GPT just up on my screen waiting. So in this case, it's much faster for me. All right.

So you should be doing it too. Find the best Chrome extension. Go through, write prompts, try them out. Again, with your context, with your background, saves you so much time processing things in your own head. All right? Let's get going. Number three, chatting with my own conversations.

All right. This one might seem a little meta. 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-grade 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. I don't know if it's weird or cool. I'm pretty sure it's awesome. All right. So again, we're showing you behind the scenes at everyday AI. This is just, you know, this is how I use generative AI from, you know, five or six a.m. until 9 a.m.

when I'm working on everyday AI, right? So, all right. So now, let's go ahead and take a look. So this is Cast Magic. And we actually had the CEO of Cast Magic on the show before. Hey, Juan, thanks. Juan said this is gold. You know what, Juan, this show is partially because of you, right?

I hear from people, you know, Juan and others say, hey, I want to see real life use cases. So thanks, thanks for the feedback, y'all. All right. So now we're in Cast Magic. And we're going to talk about how I talk to myself, weird, right? So anyways, I'm going to go ahead and click this. So normally, literally within minutes by the time that I'm done.

So like yesterday as an example, this is our show from yesterday. So AI tools to supercharge research. So we had Ivy Steeman on the show. All right. So hey, shout out, shout out Brandon, you know, my colleague here at Everyday AI. So normally after I'm done with an interview, right, I'll stay in chat with the guests after we quote unquote hang up for, you know, five to 15 minutes. So much of the time, by the time that five to 15 minutes is done, I can go in cast magic and Brandon's already crushing it. And all of this is already in there.

There's a lot of great features. And I'm going to talk about one or two. But I'm going to show you exactly how I use this because here's search. And with a I just clicked on me. You heard me talk twice. here's the reality. And let me even tell you this, how my day went yesterday. All right. So I hung up with Avi. All right. I actually had a live stream. I went on someone else's live stream. So shout out Denise for having me on your show. Then I had a 30 minute meeting with someone. There was one or two things that I did after that as well. So by the time that I was ready to read the newsletter, if I'm being honest, I had already done so many different things that I I forgot so many of the details of my conversation, which I feel bad about. But hey, maybe you can relate, right?

If you have so many different tasks on your plate, something you did an hour ago, two hours ago, you're like, man, what happened there? All right. So here's what I did yesterday. This is literally an example. Yeah, you can already see how I've been using it if you're tuning in here live. All right. So let's go ahead and do this live.

Hopefully it works. All right. So I'm going to zoom in a little bit on this. I'm going to click generate. And as it's going live, I'm going to tell you exactly what I'm doing here. All right. So I have a prompt that I just put in that said, based on the transcript of this podcast episode, please continue writing and use exact details from the conversation of what Avi

said to complete this paragraph. So essentially what I'm doing is this program cast magic. We upload the audio file of our podcast. It transcribes at all. We have about 50 saved prompts that we test and we change and we update weekly. And so when we update the pot, when we update and put our audio in here within literally three to five minutes, not only do we have a transcription, but it runs that transcription through those 50 presaved prompts that we've already done.

And then I can also go and chat with my transcript like I'm doing here. because in this example, what happened is I started to write my newsletter. Yes, the newsletter is written by me, former journalist, real human, that bottom section where we talk about leveraging, go read it today. It's written by me, not AI. But I forget so many things, so many specific details of the conversation. Because when I'm interviewing someone, sometimes I'm also taking my own notes. Sometimes I'm putting up comments, you know, on the screen, you know, like, like this,

you know, like this and then I forget things, right? So now I'm going back and I'm talking to my to my chat, right? So I essentially started writing. Okay. So here's what I wrote yesterday in the newsletter. I said one question we all might have is this. Why do we even need to use AI in research? Well, apparently it's pretty costly to have established university minds create the research our world relies on. Avi mentioned that the average cost of producing a single research paper can be more than $500,000. Yikes. Here's the thing, y'all.

I forgot the exact details of what Avi said after that, because it was two or three hours later. I had already been doing multiple other things and I'm like, oh, crap. So what I would have to do before generative AI is I would have to go and listen to this entire interview and take notes and find that spot. It would take forever. Now it takes seconds. All right.

So I ask this chat. So I say continue writing in the same silent tone of voice and respond in bullet points with every exact detail that Abi said related to this. So now cast magic's magic chat responded to me and said, hey, here's what Avi said, Jordan. Way to not pay attention. I'm an AI. I don't sleep. Here's what you miss. And he said, oh, Avi explained that these high costs, we must get our money's worth by allowing these university minds to focus, you know, whatever.

Like, you all can read this, but it's going in and it's giving the exact details of what Avi said for that point that I'm writing about that I'm like, oh, man, this is a great 34 minute conversation we had. I forgot the exact details. Cast magic, help me remember. All right. So let's get back into it. Hey, I'm curious. Could any of you use that? I'm going to tell you how you should use it.

All right. Hey, we got a shout out to the St. was post dispatch, one of my old newspapers that I wrote out, that I wrote out. What's up, Bob? All right. So here's what you should do in cast magic or a program like it. You can also, again, do this in chat, GPT. You can train a custom GPD to do this. Take your transcript of your meeting, of your, your live stream, your webinar, whatever it is. You know, in cast magic, you can build all these prompts. So here's what I would do when we've talked about this on the show before. We spend so much of our time on meetings. For the most part,

people record meetings or you use a meeting assistant. We use Meet Geek. We're also trying out Zoom's new AI companion, which, you know, we had Zoom's chief product officer on the show last week. That was a pretty cool interview. So we use multiple AI tools whenever we have meetings, but you can take the transcript of that meeting. You can build, you know, custom prompts within cast magic. You can say, hey, give me, you know, a recap, give me, you know, a follow-up. for each and every person on this call, right? Because in transcripts, you generally have to label the people. And then it says, oh, here's what you said.

Here's what Brandon said. Here's what Sarah said. And then you can build a prompt and say, hey, give me the follow up for each and every person. Write me an email response for Brandon with his follow-ups. Write me an email response for Sarah with her follow-ups. You know, tell me what I need to update in my project manager. You know, give me details of what I should do. with my CRM, right?

Whatever it is that you normally do before or after a meeting, whatever it is, you can use a tool like Cast Magic, upload your transcript, and it can all be done for you, right? Obviously, the tools that we use day to day, today are getting better. As an example, Zoom's AI companion is getting better at this. But at least right now, you can't have those kind of pre-packaged prompts. Microsoft Teams, great at this, right? And if you're using Microsoft 365 co-pilot, that really helps.

You have to have the enterprise version for it to work with Teams. You can't do that on the $20 a month pro version. But if you have the enterprise version of Microsoft 365 teams, a lot of this work can be done for you. But if you aren't doing this already, why? All right. So this is what I want you to start doing, whether it's your own meetings or you can do something else.

Maybe, you know, there's a YouTube video that you like to, you know, watch once a week. And it's long, but you like to learn from it, right? Do this. Build a custom prompt in cast magic or in chat, GPT, that is specific to that and is specific to that. and it's specific to you and how you learn.

Spend a little of time up front. You're going to save yourself so much time. I do this sometimes when I watch long YouTube videos, right? I will do something like this before I will read it and then I will watch the video, usually at 2x. But a lot of times when I've already done kind of an AI powered summary first, I will do a much better job at retaining that information as I go along because I already have a good idea and then I can really enjoy whatever I'm watching or listening to and not having

to type so many notes. All right. So that is number three. All right. Y'all, I got to take a sip here. Yeah, what Monica said, this would be amazing for long webinars, meetings, sound halls. Absolutely. Absolutely. Harvey Castro says, thanks for this episode. Hey, is anyone else finding this valuable? If you're listening on the podcast, let me know. But if you're listening on the podcast, also, make sure to check out today's show notes. We're going to link to the live stream. There was so many good things I wanted to include, right? But unfortunately, I made the title of like five things, but there is a lot more.

So if you repost this on LinkedIn, all right? So, hey, if you're watching now, if you're listening, if you're finding value, like I said, I say this almost every day. Some episodes take five, 10, 15 hours of prep work. And we do this every day. That's a lot of time. If you're finding this valuable, please take 5, 10, 15 seconds to click that repost. Share this with your network.

You're helping them. I'm not selling anything. Help your network, click repost. If you repost this, send me a DM. I'm going to send you a short video with three bonus ways to use generative AI every single day, simple ways. So if these are helpful and I got two more, click that repost. Please, if you're listening on the podcast, click that link to our live stream and repost this. Let me know you did.

And I will send you a video with three bonus tips. Hey, Jay already said he's reposting. Thanks. Cecilia said this is fabulous. All right, let's keep it going then. All right. Number four, y'all, I have a confession to make. I use Microsoft Edge.

And I like it. You know, that might not mean anything to you, but I've been helplessly addicted to Chrome for more than a decade. I like Microsoft Edge. Here's a couple things to know. It's based on Chromium. So what that means is if you use Microsoft Edge, the new, new is, new-ish browser from Microsoft. It is based on Chrome. So all your, well, I'd say like 95% of my Chrome extensions.

I'm not exaggerating when I say this. I currently have hundreds of Chrome extensions installed. About 95% of them work on Edge because it's based on Chromium. So your bookmarks, your passwords, your Chrome extensions will all work. So I like Edge. All right. So one great thing about Edge is it has co-pilot built in. And it's free.

All right. So I like to use Edge PDF summaries for fast. facts. All right. So let me go ahead. I'm going to have to share a different screen here. So y'all give me one second because I got to find my, my edge browser here. Let's see. Where's my edge browser? There we go. Bam. All right. Let's see if this works. So now I have my edge browser up. So I'm in Microsoft Edge. A lot of people don't know this. So in the upper right hand corner, and I really encourage you to give edge a try. You'll see here at the top, here's all my or some of my Chrome extensions that I have

enabled, right? I actually have so many Chrome extensions that I have a Chrome extension that allows me, guys, I wasn't exaggerating. There's hundreds of Chrome extensions here, so many that I have a Chrome extension that helps me manage Chrome extensions. Anyways, in the upper right-hand corner, you have this little copilot ribbon. Okay? So now what I'm going to do, I have this PDF open.

So I use this PDF for my show on Tuesday, my hot take Tuesday that was, hey, if you are banning generative AI, your company's going to fail. I think it was a good show. Do you guys think it was good? Let me know. If you haven't listened to it, go listen to it. So anyways, I am on Microsoft Edge here. I click the co-pilot banner. All right.

So now I'm going to click Generate Document Summary. Okay. So what it's doing now is it is reading this PDF document. So I have this Cisco study, this 2024 data privacy benchmark. And it is a long PDF, y'all. I have not read this whole thing. Let's see, I can't even tell how many pages, 26 pages. All right.

So it's giving me a recap. And it's actually a pretty nice recap here. All right. But I have a little prompt that I'm going to go ahead and put in here. Let's see if we can get this work to work exactly. Because I actually did this when I was preparing for my show on Tuesday. All right. So now I can chat with this PDF.

Yes, again, this is something you can do and chat GPT. But for me, sometimes, especially if I'm on a laptop and I don't have my 60 inches worth of monitors in front of me, a lot of times this is faster. In the age of generative AI, you have to rethink how you work. Sometimes it's about speed and efficiency. All right. So I have a prompt here. I'm going to go ahead and run it.

And hopefully I'm going to zoom in. Oh, that just zooms in on that part. Hopefully y'all can see this. So anyways, I have a simple prompt over here. And here's what I said. I said using only this PDF as a source because often even when you are using, you are chatting with a web page co-pilot, which is, you know, GPT4 Turbo. So Microsoft Co-Pilot is essentially GPD4 Turbo with some training on top of it.

But a lot of times it'll go look for outside sources, even if you are using the page as the source. So what I have to say is using only this PDF as a source and without referencing any other sources, please find the most specific stats as it pertains to generative AI use in this document, right? Because this document talks about a lot of things. And then I say, please only reply in bullet point with exact statistics and results from the study. Every response would have a number or percentage. So I'm saying, don't waste words. Please do not waste words and keep responses factually based on the contents of this PDF only

and specifically cite a stat or number in each bullet point. response. All right. So you're going to get different results every time because I just, I literally just ran this again today. And it looks like it's, it's citing some other sources within Cisco. However, all of these stats that it replied to me, they're all in this document. And if I want to see them, I can, you know, search in here and read more or it's also citing. So co-pilot is citing this. So I can click this. So 69% of organizations, concern about that the use of generative AI could hurt their legal in IP rights. So I can click that.

And what it actually does, okay, so I was wrong here. It's actually only pulling things from this document. It has these footnotes. So I thought it was linking to external links on the Cisco website. But it's actually just going to, it's a footnote. It's a jump link, which is actually really nice. So I can click any of these cited things. And it's taking me to that exact part.

in this very long PDF, right? So I'm clicking this stat. It takes me there. I'm clicking the 69% of organizations express concern. It's taking me there. Did you guys know you can do this in Edge? Good old Edge, you know, floating there in the background. People don't know.

Edge, huge time saver. Huge time saver, all right? All right. Let's go back. Here we go. So here's how you should use this. Use it exactly what I just did. Learn more and less time by consuming more PDF summaries reliably.

All right. Here's the thing. Even within chat GPT, depending on what GBT you are using, or even if you are uploading a document into the default mode of chat GBT's GPT4, all right? Because chat GPD can do this by default. However, you have to worry about context length, memory. All right?

So it's actually eating up a lot of chat. these memory. So if you are using as an example, our prime prompt polish method and you are uploading a very long PDF document, you're going to be eating into that memory and it's going to start to forget things. That's one of the reasons why I like to use Edge. It's quick, it's easy. It's literally two clicks and you're going and you don't have to worry about, oh, if I have this as part of a longer conversation, is it going to start to forget things that I've spent time training it on? No, it won't. All right.

Last, hey, Juan says it's a gem. Thanks, Juan. Juan's asking how where do you keep your prompts that you most commonly use? Good question. Juan, I don't use prompts. Weird, right? Some days I spend eight to ten hours in large language models, depending on the day. I don't use prepackaged prompts for the most part, right?

Like, as an example, the one that I showed you all to begin with, the, the AI news summary, for that one, I just use the exact same prompt. So that chat that I have within chat GPT, my version I used, I showed you guys how to do it from scratch, my version that I use and that our team uses has been trained, right? But we only use that chat for that exact purpose. And we re-use that prompt, you know, once or twice a day. So for that one, technically I am reusing that prompt. But if you use chat GPT correctly, which is you train one in each.

new chat for only one specific skill set. For that one, after I've trained it, that's the hard work. All I have to do then is just copy and paste that prompt. That's right there. So it doesn't get lost. So if you use, if you organize your chat, GPT or other accounts in the right way, if you train it, you know, if you go through prime prompt polish, you know, shout out. If you took our updated class, we have another one coming today, our refined Q method of priming.

all you got to do is just copy and paste that that prompt if it's something that you've reused daily. But for the most part, if you use chat chitit or other large language models correctly, you're not using the same chat for 50 different things. You're training one chat on one specific skill set. And then by the time you go in there, you don't have to have a prompt library. The other way, Juan, that you asked, you know, when I showed you guys, voila as an example, that one, you do save prompts because it works in a little different way.

and then you can save five by folders or, you know, you can star prompts to be your favorite and so that way they always show up when you're on a certain website. All right. Ooh, this has been a long one, y'all. Last but not least, we'll go quick on this one. So using perplexity for hallucination-free research. All right. Let's do it.

We're doing it live, y'all. We're doing it live. Let's jump into perplexity, shall we? Give me a second here. I have to stop my screen sharing. There we go. We're jumping into perplexity, y'all. If you don't know perplexity, get to know it.

We're going to have someone, hopefully someone from the perplexity's team at some point soon on the show. We'd love to get them on. But if you don't know, let's go ahead and show you. So I have a simple prompt here. This is literally a prompt that I used yesterday for our newsletter. Okay, so we had Avi on. We were talking about AI powered research.

This is so meta, y'all. I am talking about AI-powered research after the fact, referencing AI-powered research by using an AI-powered research tool. My brain hurts. All right. So I'm going to zoom in here. Hopefully everyone can see this okay on the live stream. Again, if you're on the podcast, make sure to check out your show notes and come back and watch us. So all I'm doing, so in our newsletter, we always share a resource.

So at the bottom, written by me, this saves me a lot of time researching. Okay. So we have a, you know, usually three takeaways from our show and then it says try this. And usually it's another resource. Maybe it's a tool. Maybe it's a study. Maybe it's a recent news article, et cetera. So I find a lot of them this way. So I said in perplexity, I said, how much do academic research papers cost taxpayers? So as a reminder, I had that section because Avi said, hey, $500,000. But because of time, I didn't have time to go back to Avi and be like, hey, Avi, you know, give me some examples of, you know, more on that, right? So instead, I'm going to perplexity. All right. So here's what perplexity is and what it does. So my little prompt I ran, very simple. I said, how much do academic research papers cost taxpayers?

Please give me a variety of exact estimates, specifically mentioning the year each estimate was created. Okay. So now here's what perplexity does. It goes through extremely quickly. And it's essentially reading. So it says 20 sources. For the most part, this is the default if you're using perplexity. Perplexity is free to use and you get five of these kind of more powerful co-pilot searches, I believe every three hours. Okay. Or if you pay $20 a month, you get like unlimited. Okay. So now what I'm doing here is here's the 20 different sources.

Think of how much time this saved me, right? So perplexity literally went in, read these 20 sources. And it's giving me short. little summaries, footnotes from all of these sources, right? So in this case, I don't have to have a lot of back and forth. I was just asking for a resource, right? So I can go down here. And again, generative AI is generative. You know, I'm noticing that it's not giving me the exact same responses yesterday, which is to be assumed, right? But yesterday, when I ran this exact same prompt. It gave me a research study that showed specifically the impact. So I could go in here, look at these articles. You know, I can click it. I can go in and say, hey, is this one right here a good resource for our audience to read? Right. So again, we normally break each and every podcast

down into three kind of important topics and we give you additional reading if you're interested. So that's a way that we find a lot of those articles very quickly because otherwise, I would have to do maybe three, four, five, ten Google searches and to find something extremely relevant. This is much faster. All right. So that is perplexity. If you haven't used it, I recommend you use it like yesterday.

All right. So here's how you should use it. Hey, this is number five. Don't worry. I always like to put one error in so you know that this is made by a human, All right, so swap out your normal. Go do this. Swap out your normal research queries from Google or co-pilot as an example and use perplexity.

All right. If you're not using perplexity multiple times a day, you are wasting so much time. Again, rethink how you work. How many times a day do you go and Google something? If you're a knowledge worker, if you're sitting in front of a computer, drawing on your expertise, creating pitches, creating presentations, responding to emails, doing research after a meeting, right? Why are you not using perplexity? Why?

It doesn't make sense to me. I use Google hardly ever. Until their new search generative experiences is a little better, until it's rolled out across all workspace accounts, which it's not. So Google's SGE works on my, you know, personal Gmail, right? But it doesn't work on. our company's workspace,

Google workspace accounts, right? So that's a bummer. So anyways, I use perplexity all the time. I barely use Google search anymore, right? That's one of the reasons why in our AI news that Google's like, oh, we have to start figuring out ways to monetize more because they know that they're

losing traffic to perplexity. They know they're losing traffic to, to open AI and chat GPT. You know, Open AI and chat GPT just did two big SEO plays earlier this week. which we covered in our newsletter. So that's what you should be doing. Use perplexity more.

Oh, man, this was a long one, y'all. I hope this was helpful. If so, please just repost this episode. I'll send you a video. If these five tips were helpful, imagine if you even just put two or three of these tips into practice for yourself every single day. Remember that study from McKinsey Digital

that said 60 to 70% of our time will be automated with generative AI? Do you want to win 60 to 70% of your time? That is an understatement, y'all. It is easily 80 to 90%. If you are a knowledge worker, if you do things correctly, again, I showed you low hanging fruit today. Very simple.

80 to 90% very easily. If you are a knowledge worker, constantly reading, constantly writing, generative AI will easily save you 80% of your time if you put it into practice. relearn the way you work. Repost this episode. I'll send you a video with three bonus tips. Hopefully by, I don't know, next week, right? I'll give everyone a little bit of time, especially if you're on the podcast.

I'll send this to you, you know, by next week. Just repost this if it's helpful. It helps us grow. All right. Speaking of listening, make sure to tune in tomorrow when we talk about accelerating your generative AI journey with AWS with Shrutica Parker, who, is a product marketing AI lead at Amazon Web Services. So thank you for tuning in.

If you haven't already, go to your EverydayAI.com. Sign it for that free daily newsletter. All right. Today's, you know, section, leverage section is obviously going to be five points, not three. But we're going to break down today's episode. There's going to be way more AI news in the newsletter. We always have a section called Fresh Fines.

If you need to keep up, you need to survive and thrive. new age of generative AI. You do that with our newsletter. It is free. It is written by a human. That's me. Speaking of that, I'm going to go write this. Thanks for tuning in. We hope to see you back tomorrow and every day for more everyday AI. Thanks, y'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.adobie.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.