Everyday AI Podcast – An AI and ChatGPT Podcast - EP 302: 5 Laws for Success in the AI Era

Episode Date: June 26, 2024

AI implementation is tricky. If only there were some simple, easy to follow steps to understand it all. Oh wait... that's what we're going to be breaking down with Isar Metis, CEO of Multiplai. Newsletter: Sign up for our free daily newsletterMore on this Episode: Episode PageJoin the discussion: Ask Jordan and Isar questions on AIRelated Episodes: Ep 126: Real Business Use Cases for AIEp 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 LinkedInTopics Covered in This Episode:1. Law of Continuously Improving Compromise2. Importance of Prompt Engineering and AI Tool Testing3. Scalability and Efficiency with AI4. Application of AI in BusinessTimestamps:01:10 OpenAI delays voice mode, Anthropic introduces projects.04:45 About Isar and Multiplai06:51 Insights on AI for business growth summarized.09:13 AI systems streamline customer service, benefitting companies.12:51 Reconsider productivity and efficiency in age of AI.16:18 AI tools enhance design and efficiency without coding.20:23 Advocating for smaller language models in AI.21:49 Great advice on data and optimization options.27:32 Regularly use various image generation tools for testing.28:16 Run up to 6 large models, compare use cases.33:57 Companies should prioritize proper processes and quality.36:50 Strategic and HR evaluations for AI impact.37:37 Company committees prioritize, implement and evaluate solutions.Keywords:Isar Meitis, AI tools, presentations, graphic design, proprietary data, AI policy, improving compromise, adapting to new technology, implementing AI, prompt engineering, testing AI tools, scalability, marketing, sales, operational processes, service scale, home automation industry, bottlenecks, measurable wins, strategic evaluations, HR evaluations, AI community, leveraging AI podcast, everyday AI, 5 laws for success in AI era, OpenAI, ChatGPT, Multiply, customer service, transitioning professions to skills.Send 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. If only there were simple rules to follow when implementing generative AI, you know, like a step to step guide or, you know, five laws for success.

Oh, wait, that's today. That's exactly what we're going to be talking about today on Everyday AI. The Five Laws for Success in the AI era. So what's going on, y'all? My name is Jordan Wilson. And I am the host of Everyday AI. and the show, well, it's for all of us. For me, it's for you.

It's for anyone looking to grow their company and their career using generative AI. We do this every single day, bringing you a live stream, podcast, and free daily newsletter, helping us all grow our companies and grow our careers with generative AI. So if that sounds like you, and if you haven't already, please make sure you go to your everyday AI.com. Sign up for the free daily newsletter. We recap our show every single day, as well as go over all of the different AI news and what's happening in the world because there's always a lot. So speaking of that, let's get started with

the AI news for today. Well, Toys R Us is kind of back. Well, they're back in the headlines at least and not because of their ongoing bankruptcy battles, but because of their use of AI. So the origin of Toys R Us has been reimagined through an AI generated brand film. And Toys R Us just released an AI generated short film, which was the first at the Cannes Film Festival. So the brand film Toys, the origin of Toys R Us marks the first ever creation using Open AIs Texted Video Tool, SORA, that debuted at the Cannes Lions Festival. So it was created by Native Foreign, and the film showcased the genesis of Toys R Us in a dream with founder Charles Lazarus and Jeffrey the Draft.

So Sora is obviously an AI text of video tool from OpenAI that has not been publicly, made publicly available yet, and can produce up to one-minute video clips without dialogues with a simple text prompt. So the film is actually catching a lot of flack online and on Twitter. But I actually don't think it's that bad, right? I think it's actually a good use case. Maybe, I don't know, Toys R Us didn't have the budget to, you know, produce a super high-end film and instead tapped into SORA. So, yeah, I'm curious if any of our live stream viewers caught that in what they think of it, but we'll be sharing it in our newsletter. All right, more AI news. OpenAI is reportedly shutting down in China. So OpenAI is enforcing a

policy to block users in China from accessing its artificial intelligence software starting in July. So Chinese companies like Alibaba and Tencent are encouraging developers to switch to their products. So Open AI is expanding its existing policy to block Chinese users from accessing, from accessing their AI software. So it was actually screenshots of the memos that was sent to Chinese developers that were posted on social media and kind of set this news in motion. So critics are, though, saying that this may lead to a concern for the global development and collaboration of AI. And other quick Open AI news, well, Open AI's advanced voice mode for chat GPT, which can

understand and respond with emotions and nonverbal cues, has been delayed due to remaining issues and internal checks. However, the video and screen sharing capabilities of the desktop app is still going to be rolled out and Open AI said on Twitter that it needs about one more month for the voice technology. And while Open AI is slowing down, Anthropic and Claude are speeding up. So Anthropic has just unveiled a powerful new feature for its Claude AI platform called projects that aims to enhance teamwork and collaboration with AI by providing a central hub for knowledge and insights.

So this is very similar to Chat GPT's custom GPTs feature, but it does have Claude's 200,000 token context window, which allows for more precise and relevant help from McClot. So make sure to check out our newsletter. So go to your everyday AI.com for more on those stories and a ton more. All right, but you didn't tune in for that. You tuned in to hear about the five laws for success in the AI area. So in the AI era. So I'm super excited to talk about this today because this is something we're always getting

questions on. And, you know, it's always, you're always trying to. figure it out, right? Figure out how can I actually implement generative AI. Now that we all know, there's no denying how powerful it can be for your business. But now we have to talk about how we can implement it. What are the do's and the don'ts? What are the rights and the wrongs? So it's not just me today. I'm very excited to bring on our guests for today. Isar Madis, who is the CEO of Multiply. Isar, thank you so much for joining the Everyday AI show. Thank you. I'm really excited to be here. Hey. And if you've listened before, we've had Isar on once before, he was so good. We

bring him back. He also has his own great podcast, which he had me on once or twice before. But Isar, maybe tell everyone a little bit more about what you do at Multiply. Yeah, so at Multiply, we do AI education and consulting. So we teach courses, either private to specific organizations or open to the public. And we do consulting to specific companies on how to actually implement the stuff that we teach at the course. But the goal is to help people through this transformation that or that is going through more or less any industry today through a lot of free education, kind of like what you do, through a podcast and newsletter and all the good stuff, as well as different paid channels, depending on people's

needs are. Yeah, absolutely. And one thing I love about the AI space, Isar, is it's very like, collaborative, right? Oh, yeah. Like, you know, like, I love that you're coming in, you know, on here and sharing some of your great insights. But, you know, let's just, let's just start with it, right? And let's just get straight into these, these five different laws, Esar. And, you know, let's, Let's start with number one. What is the first, I don't know if it's the most important, but what is the first, you know, law to kind of get started? Let me say one thing before that. Like, these laws continuously evolved.

Like I started with it. I talk on a lot of stages. So I get to talk to a lot of CEOs and business leaders after I get to talk. And then I teach the courses. So I get talked to a lot of people in positions in leadership. And then through the consulting. So I just meet with a lot of real companies who have real. questions and these kind of evolved through the last year and a half.

And it actually was four laws until a couple of weeks ago and they're like, oh my God, this like, so you, one of the first are getting law number five. But they're not in specific order or specific level of importance. They're just all insights of lenses through which you need to look at AI in order to help you grow the business. And so based on your request, let's just get started. So the first law is stop thinking efficiency. and start thinking outcome. And it's probably the longest one as far as how many words it has in it.

But what does it mean stop thinking efficiency and start thinking outcome? As business people, we're trained through years to think through processes, right? So everything we do, we taught in business school or through actual life that there has to be a process to it and it has to be well defined so everybody can follow it so we can scale. like all these good, amazing books was written about it, right? But so let's take an example. Let's take customer service. How does customer service works? Well, first thing, you have some kind of an intake way.

This could be an email. This could be a ticketing platform. This could be a call. This could be a bunter on social media. This could be like multiple ways to initiate a customer service process. Then somebody has to go through all those intakes and categorize them. Is this a technical question?

Is this a conceptual question? Is this a financial question? Is this? And then you have to prioritize them. So after we have them in categories, somebody needs to say, well, this is high priority. It's a big client. They're bringing us $20 million a year.

This is just a guy that's asking a general question. So you need to prioritize them. Then you need to assign them to the right people, right? So the right person. And then these people have to try to solve the problem in whatever platform is relevant. And then you need to document it. So in the future, it will be easier to solve.

And then it starts all over again. And when you go to people in that field of customer service, you tell them, okay, we need to improve our customer service because we don't have enough budget to increase the team, but we got to do this better. So what do they do? They're going to say, okay, let's try to improve the intake by not allowing people to call in or we improve the IVR. So click one for this, click two for that.

We're going to improve this. We're going to improve our ticketing system. So it's more specific on how people open tickets. So that's going to be step one. And then that's going to give us a 5% increase in efficiency. And then step two, we're going to define a better way on how to prioritize and so on and so forth. They go through every step of the process, try to find small efficiencies in the process. The reality is today right now, as we speak, there are fully capable AI customer service systems

that can do the intake priorities. prioritize the thing, provide answer 365 days a year, 24-7 in multiple languages, connected to your CRM, to your ERP, to your financial system, and they can do everything a customer service agent can do, better, faster, cheaper, including large companies like Klarna that has done this at scale and have cut off 700 jobs, and I'm not suggesting that's what you need to do, but they probably had 20,000 of those people doing customer service. And they were able to provide customer service, at three minutes average time to close the task versus 13 minutes while getting the same score for their level of customer service.

So basically what I'm saying is they've circumvented the entire seven steps that we talked about because the goal, going back to start thinking outcome, the goal of customer service is not a better ticketing system. It's not a better IVR. it's not a better process. The goal is happy customers. If AI can deliver happy customers, faster, better, cheaper,

forget about the process. Now, one more thing, and then I'll let you ask you a question. In some cases, you cannot do the full jump. But in two jumps, you can go 70% of the way. So what I tell people is, again, forget about your existing process in literally everything in your business.

Think about what, is the outcome that you're trying to achieve and then look for AI tools that gets you most of the way there, either in one jump or in a few jumps, that will save you a lot of the steps that you may not need to do anymore. Because if you do it the other way around, you're going to miss huge savings because you're going to try to improve little things in your process. Yeah, Isar, I think this one is super important and it's worth diving into a little more. You know, one thing that I tell companies that we work with is you have to unlearn good habits. because we've especially since, you know, the SaaS scene and, you know, all this software

has that has become available over the last couple of decades has made us more efficient. It has made us more productive. And that's not necessarily now in the age of generative AI. That's not the best way to always do things, right? Just because they've worked or just because, you know, a certain way, right? Like I find myself personally very efficient and very productive in how I work. but that doesn't necessarily mean going forward that that's the process I should be relying on, right? If there are generative AI systems or large language models that can do pieces of it better.

How can business leaders, you know, kind of tweak their thinking? Because, yeah, I think for so long, we have been, you know, it's been pounded in us like, hey, if you're efficient, if you're productive, you will be successful. And sometimes I think what I see now is people just being still being very productive, being very efficient, but doing it kind of the quote unquote old way, how can people tweak that way of thinking? I think the first thing is mindset, right? It's exactly these things like look for knowledge on how to do it differently,

look what other people are doing, follow people like you or me on LinkedIn, TikTok, Instagram, wherever, people who share actual use cases, who can show you what's possible. And the other thing is start a committee in your business, start a group of people, there's going to be your, AI committee that can brainstorm and then set these guidelines. These are the rules like these lenses that I'm talking about right now. This is how we're going to evaluate everything in our business.

And then even if you're not thinking about it, once you have four or five or six or 12 people, depending how big your company is in your committee, somebody will come up with the idea and say, hey, you know, I saw this thing online. I think we don't have to do these five steps. Or I think maybe the five reports that we're looking at every week is not the best way to get the information. that we need to get to make decisions. All right. And hey, that's great and very excited to jump into number two.

But just as a reminder, for our live stream audience, joining us. If you do have questions, please get them in now. We'll probably have a little bit of time at the end to go over anyway. So, Esar, now that we have the first law, you know, to kind of think more outcome and not necessarily efficiency, what is the second AI law for success? So the second law is from profession to skills. And that's a cool one that I really, really like because I'm a very big example of that. So if you go back, by the way, to the 17th century, people had the title of computers

because there were very few people who could compute. Now, which basically means if there were business cards in the 17th century, which I don't think they were. But if there were, some people had computer written on their business card. I don't know many people have computer written on the business card right now. Another example is typist. My next door neighbor when I was a kid, still good friends or my family,

was a typist. That's what she did for her entire career. And again, I don't know a lot of people who are typists right now. So all of these things were professions that became skills, and they became skills through the implementation of technology.

So technology enables us to take something that was a profession that you actually went to school for and made it available to everyone. And this has now with AI I've been dramatically accelerated. So when I talk about myself, as an example, I've never studied graphic design.

I've never studied how to write code. I've never written a line of code in my life. I'm a really, really bad accountant, and I've never studied that. However, now I have amazing tools that allow me to create the designs of everything that I do, whether it's presentations that I'm giving on stages, presentations to clients,

courses that I teach, etc, et cetera, et cetera, like all the graphic design for that, 100% of it is generated with AI without going to third-party graphic designers that I used to spend a lot of time and money in getting their help. And it's not that they became bad. It's just a scale that I was able to acquire using AI. As I mentioned, I don't have a clue how to write code. I don't understand code, but I don't need to.

And yet now I have different codes that are running in different things that are making my work more efficient because I'm using Chachipiti and Claude, and I actually use them back and forth to troubleshoot its other issues in the code to get it to work and do the things that I needed to do. Now, will that enable me to program the next FIFA game? No, probably not, but does it allow me to, or at least not in the near future, right? But it allows me to do a lot of things more efficiently in my business in ways that were just not possible to me before.

Now, the same thing goes to, you talked about SORA, the same thing is going to go for actors and screenwriters and filmmakers and lawyers and paralegals and home automation experts and so on and so forth, right? Because there's going to be AI tools that's going to learn that specific topic and will allow anybody to do this at a professional level. Yeah. And so our next rule, Isar, is about how we can win. And everyone wants to win with generative AI.

So what is rule number three for our AI laws for success? Awesome. So rule number three is that there's two ways to win in AI in the AI era. One is having and leveraging your proprietary data. So if you can train models, and I'll talk about this in a minute because I can train, I cannot train models. I'm not Google. But you can and we're going to talk about this in a minute.

But if you can train models on your proprietary data, you by definition can get insights that your competition. do not have because they don't have the data, which means you can make better, more data-driven, educated decisions when your competition doesn't have access to that. So there's two things that people say. One is what I said. Like, I don't know how to train models. Well, that concept of training models sounds really complicated, but the reality is with even just what you just said this morning, if you build a GPT and upload five documents to it, you just train the model. It's called RAG and it doesn't matter what it means because I don't want to

confuse people, but it basically means you just train a model on a specific piece of knowledge from your company. So now going back to, again, your news from this morning, if you have a 200,000 token context window, you can probably upload 20 different documents, which could be your winning proposals and two examples of bad proposals. These could be good ways to do customer service, bad ways to do customer service. This could be HR, like all your HR documents that everybody goes to HR. to waste their time to ask them how to get a PTO or apply for maternity leave or whatever

people come to H.R. And so you can train models on your propriety data very easily. Like you don't need any special IT capabilities. Literally just build a GPD connected to things. And obviously there's more advanced ways to do that. They're still within the reach of most companies. So proprietary data is one. By the way, the other thing people ask is I don't have proprietary data.

Again, I'm not Google, like what kind of proprietary data I have. So every company has proposals and customer service and emails back and forth with clients and success stories of stuff that they've done online that work well from a marketing perspective. Like each and every one of those is a piece of data that you can train models on in order to create more of the successful stuff and avoid the not successful stuff. 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, Illustrated Premiere, light, light, 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. at firefly.adobie.com. That's such great advice, right? Even as we talk about, you know, GPTs from, you know,

chat GPT or the new projects kind of feature from Anthropic, right? And I've said this on the show a lot. I think that the future of large language models is actually small language models or this concept of working with multiple smaller GPTs or multiple smaller you know, now projects. I wish I wish these companies had like better naming mechanisms. Oh my God. You'd say projects and GPTs and it's like, what the heck does that mean? But, you know, I think that is the future of how we are going to be working. So I like that we can get a taste of this right now in, you know, these consumer facing tools, you know, in chat GBT and

in Anthropics Claude. One thing to note there, which, you know, I'm sure like Esar and I talk about this all the time is like when it comes to your company's data, always make sure before you run and you know, go upload that. Make sure you read your company's AI policy handbook first or, you know, talk to someone, you know, make sure, you know, that you have a good game plan when it comes to what files you should or shouldn't upload, you know, depending on your plan, whether you're on a, you know, chat GPT Plus or enterprise plan, there's different data sharing. But it's, I think it's important to, you know, keep that in mind. Isar, what is our fourth kind of AI or law for success in the AI area. So first of all, great advice about the data. Don't upload data to any

of these platforms before you know whether you should or you shouldn't do that. So that's a great advice. But still in this law, I said there's two ways to win. One is data proprietary data. The other is optimization. So you can optimize your processes within your business, whether it's marketing, sales, design, product, QA, HR, finance, like each and every one of those has multiple ways where you can optimize them with AI in order to be a lot more efficient in them. So let's say everything is vanilla and nobody has proprietary data and you can't have any real edge above your competition in your field. You're in a highly commoditized world. Then you can now do things 20, 30 percent more efficient than your competition, meaning by definition

you can now lower your rates by 15% and still make 15% more money. So your customers are going to be happier. You'll be able to deliver better work faster and you will make more money and they will pay you less. So literally everybody wins. So proprietary data is one. Optimization is the other. And obviously, if you can do both, then you win on both sides of this equation. So that's this law.

Now to the next one, that's the new law that I told you that I added rule number five. And that's because a lot of people are disappointed when they start using AI. I'm like, oh, this is bullshit. Like it doesn't work. I can't get consistent answers, blah, blah, blah. And it's true. Like, I'm not disqualifying this. So the law is called the law of continuously improving compromise.

And what it basically means is that, yes, these systems are not perfect. But they're good enough for a lot of things once you figure them out. And the problem is that people get into this and they try one or two things and they don't work exactly the way they want it. Or they try the same thing three times and they get three different results. I'm like, okay, this is bullshit. I can't use this in my business. And they stop. And the reality is if you invest the time and the resources to figure out one use case, two use case, three use cases, and you end up with 30 use cases, each and every one of these use cases gets you an efficiency in something.

And if you don't do that, if you say, well, I'm going to wait till this thing is perfect. I'm not going to compromise. What's going to happen is, A, everybody else is going to figure out these small efficiencies along the way and you're going to left behind. But even at the day, it's perfect. You will still have to figure out how to implement this. You will need to figure out because the tech is a small aspect of this in a business. There's processes. there's training, there is do's and don'ts, there's regulations.

Like there's a lot of stuff beyond, oh, I'm going to flip the switch and buy everybody the licenses. That's the smallest part of all of this. Figuring out how to use it properly in a business context when there's teams, there's people, there's training, there's divisions, department. Like all of this has to come into play when you start implementing these things. So if you start now and you're saying, yes, I know I need to compromise. And I call it continuously improving compromise because, as you said, Anthropi came out with Claude 3, three months ago.

And now we had Claude 3.5 that can do a lot more stuff and faster and better and a higher accuracy. So you continuously need to compromise on less things. But if you start now, then you start gaining benefits. You learn how to use these platforms. You find ways to implement them today, the way they are today. And they're good enough today for many, many different. things, not everything, but for many different things. And so start today, improve as these platforms improve, and don't wait for it to be bulletproof

because then you're going to be left behind. Yeah, I think that's important. And, you know, two small pieces there, Issa, is something that we see all the time around that is number one. You have to know the basics of prompt engineering, right? I'm not saying that you have to become an expert, but the future of work is generative AI. You have to understand the basics of how models work and how to get the most out of them. That is prompt engineering. So if you're just trying to, you know, trying to find someone online and

trying to find their prompt that they use and, you know, get a copy and paste mega prompt, that's not how large, that's not how models work. That's literally not how they work. So that's important to keep in mind. Another thing is you shouldn't be expecting perfection, right? Like I love Isar that you said it is continuously improving because yes, on a week to week basis, you need to be looking at your processes, looking how you're using generative AI, and you have to be continuously improving it. You know, I say, you know, maybe if you do it correctly, you're going to get 80% of the way there and only 20% of the time.

It is not a, you know, copy, copy and paste 100% automation, 100% of the time. So I love the concept of having to continually improve it. All right. We are near the end. Isar, what is our last law for AI success? Just before I dive into that, I want to pay. piggyback on the thing you just said on on the tools continuously improving one of of things that I get asked a lot and I'm sure you do as well like okay so which tool do I use

do I use chatypT do I use Claude there's Gemini do I use whatever and the reality is different tools work in different ways for different use cases so I use all of them every single day I use Claude every single day I use perplexity every single day I use ChachypT almost every single day, I use different image generation tools, whether it's mid-journey or others. And the reason is, I use them all the time, and I test them against each other all the time. And I do that to figure out which one is doing this particular use case better. I'll give a small little tip on what's the easiest way to do that, at least on the chat side. I use a Chrome extension that is called chat hub.

So there's a lot of chat hubs if you Google them, but this one's just a Chrome extension. So you go to the Chrome extension store and you get it. And I got it early on. So there was like a lifetime deal for like $30, which I don't think exists anymore. But there's a free version of it as well. But what it allows me to do is it allows you to run up to six.

I always use four large language models in a single chat. So you have all four of them on the screen and you chat on the bottom in like your entry. and you see all four models running at the same time. And this is an amazing way to compare specific use cases. Okay, so here's my problem, whether it's short or long or whatever, you put it in and you see four different models running at the same time, and you can then know for this particular use case, after you do it two or three times,

because once is not necessarily going to give you the right answer, which one does it better, and then I go to the actual tool and use the actual tool. So that's just a tip on how to do what you mentioned. Last law. So the last law is, A, enables infinite scalability. And like, okay, there's no such thing as infinite scalability. True.

There is no such thing as infinite scalability, but it's pretty damn close. So again, let's think about a business. A business, every business starts with marketing. So people need to know that you exist and your services and products. And then there's sales. So, okay, people need to engage with your company one way or another to buy your products and services. Then there's some kind of operation where you deliver the goods or the services.

that you promised, then there's customer service, and then there's all the back-end stuff, right? There's finance and HR and all the other things. So that's how the business works. And every business has bottlenecks. So if we go step by step, like, okay, this is, we're limited with the amount of marketing we can do. You hear that with every company you talk to. I work with companies of five people and of 150,000 people. And they all have, well, we're really short with our marketing resources. You've got to understand this. I'm like, I get it. Everybody doesn't have enough marketing resource. But is it true?

Now, you can literally, two people with the right AI tools can generate SEO optimized blog posts probably 30 a week when they could do two before. They could generate social media content basically as much as they want. So if you don't want to post more than once a day on every platform, you can generate optimized content for every platform. So not the same thing, but actually optimized for TikTok versus Instagram versus YouTube versus LinkedIn, different content with two people for every day of the week

without running into bandwidth issues. And the same thing for brochures for your next trade show. Like every one of these things, you can generate significantly faster at significantly a higher scale. So that solves the marketing thing. So, well, then I don't have enough salespeople. I just don't. Like, okay, if I now have 30x, the amount of leads, even today, we can't handle the number

of leads we have. And so how do I do that? I'm like, okay, so some of the sales process AI can do. It can analyze the initial responses on social media initially or the initial emails for request. It can answer today, right now. I'm not talking about somewhere in the future. That exists right now.

Or at least help prioritize them of what people. should handle first versus second to increase the efficiency dramatically. The scalability, so the same number of salespeople can now handle 3x, 5x, 10x, depending on how good you become in this, the number of leads. Like, okay, awesome, but I don't have the people to actually do the thing. Like whatever, so if it's a product, it's not a big deal. Okay, now that you just have more people buying the product. But if it's a service, then I don't have enough people doing the service.

Well, the same thing. The service can be scaled with using different AI tools. So one person doing the service can now potentially support three people instead of one. So now you can scale without hiring people on the sales team. But there are stuff where you have to hire people. So if you are, a lot of my clients are in the home automation industry. They need installers. They need people actually going to home installing speakers and TVs and smart lighting and all the other stuff.

So there is no, I don't need more people. You need the people. Well, you can hire and train people a lot faster using AI. You can write better, more accurate job description. You can analyze the places where you posted them in a much more accurate way to see where you're getting the better people. You can help it write interview questions and help you during the interview process or evaluating your resumes that you're getting. Like literally every step, the training for sure, like with a generation of videos and training and stuff that can be done in minutes instead of hours. So all of these things as far as hiring people and training people becomes a lot of that.

So literally every aspect of the business, AI enables you a lot more scalability. And the trick is look for your bottlenecks first. Start with the things that are limiting your growth right now and say, okay, these are the three main things that stop us from growing. And look for AI solutions where you can grow 5%. Forget about the 50%, 200%, 500%. 5%. But if that's your bottleneck, now your company. has 5% more throughput. So that's how you need to look through this lens.

Yeah. And I love, you know, starting at the bottleneck, right? I think that's huge, especially when it comes to AI implementation or if this is your first, right? If you're listening and this is your company's first big AI initiative, right? Sometimes people scope out like year-long projects. It's like, no, like you need to find a short, quick, measurable win. And I love starting at where you bottleneck.

One other thing, Isar that I think is important to talk about, you know, when we talk about content generation and infinite scalability, yes, absolutely. I think one mistake companies are making, though, is they don't go through the proper processes first, right? And they don't, you know, learn the, you know, 101 of, you know, prompt engineering. They don't learn how to actually create good content with generative AI. And then instead they just, you know, have this almost like robotic or lower quality content at scale. So I think it's important, right? Esar has fantastic advice here, but you also have to

figure out. You have to make sure the quality is, you know, high first before you just put that thing, you know, on scale, you know, all across your organization. So Esar, we talked about so much in today's episode, right? So I'm going to do a quick recap here and then ask you for your one kind of best piece of advice to put these into practice. So number one is to think outcome. And I'm, I'm summarizing here. We'll get them all, you know, as they should be in today's newsletter. But number one is to think outcome, not efficiency. Number two is kind of going from professions to skills.

Number three is there's two ways to win, both with proprietary data and optimization. Four is this law of continually improving, kind of continually improving with compromise. And then number five is AI enables infinite scalability. So, ESAR, so much great information packed into a very short amount of time. How can people actually make use of these five laws of success? How can they get them, you know, working for them today? So the first thing is to actually take action, right? So a lot of people are afraid of this or too confused or terrified or just not paying attention.

Like there's different excuses. But start doing stuff, right? The only way you learn, like you can, I will add to that. Like, people who are listening to this are already doing the first good step, which is to educate themselves from people who are actually doing this. So if you're listening to this podcast, you're already doing one step in the right direction. So follow the right people to get the information. But step two is you've got to start taking action because you can listen to Jordan or myself or a lot of other amazing AI experts out there. If you don't actually start, you're not going to gain any benefits from it, even if you become really, really smart.

And even the stuff that we say until you start experimenting with your needs, in your environment, with your niche, with your technology. technology with your knowledge, you won't really gain the level of understanding you need in order to actually implement this. The second thing, as I mentioned, don't do this alone. Like, find a group of people, preferably within your organization, or both. You can have an external support group, but that can help you brainstorm and implement and test and evaluate different use cases. And you really need to start with three different evaluations, and I'm going to name them very, very quick. We spend a lot of time on this when we actually work with companies. But one is a strategic evaluation. What's going to change in my industry, in my business, in my niche because of AI?

Like, what will my clients not be interested in two years and what new things will be able to offer them in two years? So the strategic evaluation. The second one is HR evaluation, a skills gap analysis. Like with all these new AI capabilities, as I talked about from professional skills, what gaps do we have as far as skills in the company? What do we need to hire for? What do we need to train for and how do we do both these things. And then the third one is the bottom up approach. So we talked about the top down the strategic side. The bottom up thing is bottlenecks. Like what small things, low hanging fruits we can change in our company today. Like with a one hour of investment, we can save one hour a week. And there are dozens of these in every single company.

So if you put a group of people, that committee that I talked about before, from every aspect of the company, so you have a person from HR, person from sales, person from finance, person from whatever, and they sit together and they look for these things and they prioritize them. And you then look for follow people like Jordan and me and others and that provide solutions and say, okay, this sounds like something that's relevant to our business. Let's start implementing it. And you start testing and evaluating initially small without sharing the wrong data while defining rules and guidelines. like all the stuff that we talked about before. And then you go to the next one and the next one and the next one. So much great advice. And, you know, I know sometimes because, you know, having a daily podcast, you know,

I hear that advice like start every day. And like, like you have to start somewhere. And I think people might get tired of hearing that. But it's, I mean, you have to, right? Like you, you can't just continually be learning and brainstorming and, you know, committing your way to this. You actually have to start somewhere. And I love what Isar even just said there.

What one hour, one time investment can then lead to a one hour gained a week? Start that small, right? Let's build something one time that we can win back one hour a week. I love it. I love it. So much great, so much great advice there from Isar, the CEO of Multiply. Isar, thank you so much for joining the Everyday AI show a second time in sharing your insights with us. We really appreciate it. Thank you. I'll say something to the audience.

If you're listening to this on the podcast, open your podcast platform and give Jordan a five-star and give him a review for the stuff that you're learning. It sounds obvious to do a daily podcast. It's very much not. It's a lot of work. So appreciate what he's doing. Give him a review. I'm sure he'll appreciate it too. Got to love that. You know, fellow fellow podcaster, you know, putting all that good information out there. Thanks, Isar. Yeah, and go, we'll link to, you know, I was recently on Esar show. He brought together for his 100th episode, a bunch of, you know, some of the best minds

in AI. So I'll make sure to link to that in today's newsletter. So make sure to give that a listen as well. So, yeah, so I will say something about this. Just one last thing before you go, because you said earlier that, you know, the AI people is a very supportive community. So the 100 episode of the generating AI podcast, I basically sent an email out to people like Drone and say, hey, will you be on my show a week from today?

I thought, I get, you know, three, four people say yes. And like 20 people said yes. And it was absolutely, it was a good problem to have, like having 20 experts instead of five show up for your 100th episode. But thank you and thank, you know, everybody else in this industry that are very open and sharing. I appreciate that. Right. Hey, you are going to appreciate if you read today's newsletter because Esar dropped so much knowledge. I'm going to have fun going back to write this one. And obviously he's on point. He was talking about chat hub. I'm like, I love chat hub. We've

done a couple of videos on that. So we're going to be, you know, recapping every single thing that we talked about. So maybe there's too much good advice in there, you know, whether you're walking your dog or on the treadmill, whatever. It's all going to be on our newsletter. So thank you for joining us. Please go to your everyday AI.com. Get that free daily newsletter. and we'll see you back tomorrow and every day for more everyday AI. Thanks, y'all. Thank you. 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 refund. 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.