101 | The Biggest AI Online Event of the Year Part 2

0:00Welcome to part two of episode 100 of the Leveraging AI podcast. if you have not listened to part one go back one episode and listen to it first it's a big event and if you're only going to listen to part two you're only going to listen to half the amazing experts that participated and share their best tips on how to use ai in your business so go and check that out first and if you have listened to part one go and check out part two and obviously if you like this podcast in general i would really appreciate if you subscribe and share this with other people who can benefit from it. In the next few years, AI technology will change our world dramatically.

0:45Whether you are a business executive trying to catapult your business forward, or just somebody who refuses to be left behind and want to advance your career, this is the show for you. I'm your host, Isar Maitis, a serial entrepreneur and an AI enthusiast. You'll hear invaluable practical tips from innovative business leaders, AI practitioners, and some of the brightest AI minds in our world today on how you can leverage AI in ethical ways to advance your career and grow your business.

1:20And next we have Jordan Wilson. So Jordan Wilson is a successful marketing agency owner for years, And he has jumped all in into AI very early on with his Everyday AI show and newsletter. If you're not listening or being subscribed, it's one of the people here that you want to follow their stuff. He is amazing in creating AI solutions for more or less anything in his business. And he's going to share how to use company first party data with AI in order to win. As a data geek, I'm really excited. Jordan, let's go. Isar, thank you so much for having me. It's good to be back with y 'all. So let me just see if I can share my screen here.

2:03There we go. Bam. All right. First of all, the guests today have been amazing and everyone's doing a superb job of sharing different tools and techniques. So I wanted to talk about something different. I wanted to go big picture because there is a huge problem right now that's happening with data. No one's talking about it. I'm going to give you hopefully what you need to know in a couple of minutes and what your company can do about it. All right. Real quick about me. Isar did a great job, but we are an AI media company and consultancy. So we've worked with thousands of individuals on prompt engineering and AI implementation.

2:38But here's what I really wanted to talk about and what's happening. AI, first of all, sorry, y 'all, I'm a realist. AI is going to both create and take away a lot of jobs. All right. Let's just put that out there. But our roles as skilled marketers, admin, analysts, et cetera, they're going to change. And companies need to be preparing for that change now. It is great to use all these great tools and prompts, et cetera, but that's not going to help you succeed in the future. Here's what we need to know. Right now, McKinsey study, we've seen this. 60 % to 70 % of knowledge workers' time is going to be automated by generative AI.

3:15The World Economic Forum, 60 % of jobs in the West could be impacted. Goldman Sachs, 300 million jobs can be impacted. Oh, it's terrible, right? Not really. Not if you are preparing your company. All right, here is the issue. Literally no one's talking about it. ChatGPT is not new, y 'all. The GBT technology has been widely used since 2020. And as someone that has offered SEO in the past, so much of this data that is going into these large language models, these multiple trillion parameter large language models has already been created using AI. And actually studies are showing that 90 % or more of the content that we see online is going to be AI generated by next year.

3:58That is wild. It is also extremely problematic because data quality is a huge issue. Right now, smart companies have figured this out, but you even have to the point now where universities in the US are selling their students data and research because they know that large language model companies need better data, all right? So here's another thing that you should be doing right now. There's going to be a major shift on how you should be upskilling your employees and how you need to have an AI-first mindset. Here's what you gotta start doing right away. Yes, jobs are going to look much differently.

4:40If you are a knowledge worker, if you spend the majority of your day sitting in front of a computer like so many of us do, getting paid for your knowledge and expertise. That is not how you are going to be rewarded in the future. All right. It is going to be how you can leverage generative AI, both in your role and your company. So we've talked about RAG, right? Retrieve Log Metageneration now for the last year or two. And this is about how you can bring your own data first before it goes to the large language model. And this is the crux of what I want to talk about today because company success in 2024 will be how they turn their company's IP, their first party company data, their expertise into high quality data for their large language models.

5:23Whereas right now we are relying and probably over reliant. If we talk just as an example about Gemini 1.5, Claude 3 Opus, GBT 4.0, unfortunately, it is a lot of low quality data. The training is getting better and better. However, if you want your company to succeed in 2024 and beyond, you need to be thinking about what happens if generative AI implementation works at your company, right? Everyone's investing so much time, money, and resources, but what happens when it works? You need to have a plan with what happens with all of those knowledge workers who are now 60 % to 70 % more productive.

6:00And I'm telling you right now, the companies that succeed next year and beyond are going to be the ones that are putting those people to use recording and turning their first party company knowledge into data to use with large language models. Bam. Oh my God. This was awesome. I love this. And I agree with everything you said. Any offer that you are making or want to make? Yeah, we don't do this a lot. I think we've only done this like twice in 15 months, but we are going to be offering a 20 % discount on our biz edition of PPP. That is our prime prompt polish course for businesses. Normally don't do discounts, so I'll throw that in the chat, but Isar, thank you for putting this together and for having me.

6:43Awesome. Thank you so much. If you want to know more, we had Jordan on episode 90, sorry, 82 of the podcast. And next we have Celeste Amile. Celeste has been in the creative world her entire career, and she's now mastered how to use AI to put some jet fuel into her creative juices, and she's teaching other people how to do the same. She's going to share how to master content creation with GPTs without getting generic results. Celeste, so awesome to have you back. Awesome. Yeah, I'm really glad to be here again. So let me just start right away sharing in my screen. And now, just as a quick reminder, you're going to be able to have this guide available in the resources that I'm going to show you because I don't think we are going to have enough time to cover all of this.

7:34And just make sure that you can duplicate this as a template because this will be available only for a short period of time because this is just some of the things that we covered in my program where I help people to build personal brands with AI in a smarter way. This is a structure to build GPTs that are really useful for your content and not just trash. So the best structure to build your GPTs for content is just to use a rare framework. So here you can say, just act as role, follow a specific action to achieve these results, emulating my examples. And the way that you do this is, for example, hey, GPT, act as role.

8:12So for example, copywriter or social media expert, or then follow a specific action. And for this, This is an awesome hack I just discovered because I didn't understand. I didn't. Yeah, because I didn't understand why my GPTs were so unstable. Sometimes they did just handle the things that I asked. And I was like, why? I'm just telling you all the instructions. Please follow them. But the thing here is that the longer the prompt is, ChaiGPD will have a harder time to finish and to follow all everything. So if you have step one, create the hook. Step two, analyze enough information. step three just review your training data it will be more stable then you explain the results that you want to get for example providing results of good versus bad behaviors for example what specific tone of voice you wanted to use what words you don't want it to use or how long you want your content to be for example i just analyzed this gpt this this post from linkedin And I realized that none of these lines of text was longer than 60 characters long.

9:20So you can say, hey, I don't want you to give your output in more than 60 characters. And when you put all of these examples together, you can basically have a GPT or build a GPT that can do this. In my case, I wanted to create a GPT that helped me to do the format of my thoughts. So I'm here just explaining the thoughts that I have. And the GPT then, because it's trained like a copywriter, a professional copywriter, is understanding my data. It's analyzing the copywriting frameworks that I provide. And finally, it's giving me just the perfect draft of my post for LinkedIn. And this is how it works.

9:58Or other GPT, I'm just providing and generating hoops in seconds. And this is not just random data. Actually, most of the GPTs that we have in the program are built with the knowledge of Jim Edwards, who is actually one of the highest paid copywriters in the world, just because he has a really deep understanding of this. And this is just because if you have really good data to train your GPTs, you can have awesome results. and again in the guide I have I cover more of these some types of prompts that you can use like the voice of tone character control how to avoid trashy language just some of my favorite prompts that I have that you can put put in your GPTs and make them valuable and also here like in this part I love this part because I give you an example of a prompt that I have I give you I give you just a small part of the prompt but the original is actually 5 ,000 characters long And when you have these problems that are long and that really work, you can see why some people say, hey, I don't know how to create powerful content with AI.

11:01But it's just because you don't know the science that is behind the content or you don't know the science of how it looks like to think like a machine and actually explaining your thoughts to a machine that can think like that. So, yeah, this is always part of the things that I wanted to explain. Amazing. Any offers you want to make? Sure. Just for this occasion, we have a 30 % of discounts on the main program that I run where I help. I think I already said that, but we help entrepreneurs building their personal plans with AI. And yeah, I just made this guide with all the love for you. We don't have enough time for this, but if you read it, I'm sure it's going to be pretty valuable for you.

11:43Amazing. Thank you so much, Celeste. I really like what you're doing. She said in two words in the middle, yeah, I analyze data from LinkedIn. If you don't know anything about data on LinkedIn and how you can analyze it, we have an amazing episode with Celeste that goes deep into that. That's episode 88 of the Leveraging AI podcast. And next we have Carl. Yay. So Carl spent most of his career in digital marketing. He's an expert that has done this in multiple companies, and he has deep expertise in taking his technical skills together with his curiosity to create really magical stuff. And now he's doing this by helping people in the real world combine AI with what they're trying to do to create better marketing.

12:23And he's going to share with us how to become your local area's go-to AI expert. Carl, you're up. Thank you very much, Sar, for the invitation. So I know there's, mine's a little bit, my presentation is a little bit different, more on business development. Because we, over the past about four, say six months, we've been really growing. but I think the challenge is I know there's a lot of people here who are growing their consultancies, solopreneurs, growing their AI agencies. And obviously the challenge for a lot of us is, Hey, how do we actually reach the businesses that we want to reach? And then how do we get them to pay for our actual AI services?

13:04Because for a lot of companies, when we do our discovery presentations, it's brand new, right? It's not like it's easy paying for marketing services, consult business strategies. But when you have companies that are like, wait a minute, I'm going to be paying for a brand new AI service, whether it's AI automations, strategies, full implementations, custom builds, and so on. And so this tip, and we've used this quite significantly to grow our business quite a lot, is really you've got to go on a speaking tour. Now, before I share that, just to let you know, it may not work for you because our target audience is medium sized companies, non-tech, very traditional industries, construction, manufacturing, metal contracting, oil and gas, economic development.

13:52You could say tourism is in there as well. But what we figured out is that locally, and we're targeting local audience, is that we just went on a presentation tour because we found that a lot of industry associations locally, manufacturing, builder groups, construction groups, economic development, public service groups, they usually have, let's say, a weekly, biweekly, monthly, quarterly get together. and they're always looking for speakers. They're always looking for that lunchtime presentation. They're always trying to find people to fill it up. We just sent emails around, notes to their websites.

14:33And because AI is such a hot topic, they love to have us out. And so we come out, we do a discovery presentation, not so much on what AI fundamentals of AI, just actual business use cases. So we do presentations on here's some AI automations, examples. Here's video and audio AI. Here are digital assistants chatbots. Here are advanced large language model use cases. We just went to France for VivaTech. We showed them the new GPT-40 demo, like all these different things. And you're reaching like, we're reaching our target audience, but I think you can be, as you present to more and more of these groups, what we found is people just talk, Hey, zero to 60.

15:27Yeah. They came and did our press. Could you come to actually our company and do the exact same thing you did? And so we are reaching a huge, really big audience and our target audience in literally free presentation, like lunch and learn presentations to all these groups. So if it's something that you, these are your targets and something that you want to be known locally for, to being that AI expert, that's, I would start getting on some of these lists, getting those connections, and then even just emailing or going to their websites and offering just a quick presentation on AI. And I'm sure they'd love to, they'd love to have you.

16:06And that's amazing advice. I love the focus. I love the niche that you pick, you know, let's know exactly what we're competing for and let's find a way to connect with them. And I think this is an amazing advice to a lot of people here listening. Are you making any offers today or just sharing amazing content? If you are in those industries, we'd love to talk to you. We have, for us, we have a usually 30 to one hour discovery presentation and even training workshop that we offer for a lot of our target audience. So yeah, just connect with myself on LinkedIn. Awesome. Carl, this was fantastic. Carl has been one of the early guests of the podcast.

16:44So if you go all the way back to episode 11, which was probably May of last year, I assume. So over a year ago, you can find my episode with Carl. And next we have Ashley Gross. Ashley is a marketing rock star who turned in an even bigger AI rock star. It happened overnight, at least for me following her on LinkedIn. She has even held a position of AI task force lead before AI became cool to everybody. So before ChatGPT, and she's going to share with us how can AI combat misinformation, which sounds counterintuitive to me. So I'm very curious myself. Ashley, welcome to Leveraging AI again. Let's freaking go.

17:30I know security, it's super sexy, probably the sexiest topic on the agenda today. Someone's got to do it. Might as well be me. Glad you took one for the team. Thank you. No, when I put a poll out of any tutorial anyone wanted to see in the next five minutes, I got a lot of people that wanted to see how to use AI to combat misinformation, which is totally fair. Because I like pain and I like to test myself, I decided to show you as many AI tools as I possibly can in the next four minutes. I will drop them in the comments below. So all I need is your undivided attention for the next for minutes.

18:02This first one I'm going to show you is called originality.ai. I had good old perplexity create a blog about the benefits of nature. And originality.ai actually detects how many AI generated words are showing. It gives you a breakdown of the highlighted color, 90 % confident this was generated by AI, orange is 70%, yellow is 50, and so on and so forth. This was the 100 % AI generated blog that is popping up as, all right, I generated my own little blurb about nature. And we're going to scan it so that you can see the difference. Maybe.

18:41Oh, I love lives. Again, the beauty of lives. It's going to be amazing. Here we go. Here we go. Let's go. 99 % original, 1 % AI detected. However, it's not telling me what part it thinks is AI detected. And 99 % is good enough for me because I genuinely wrote this. If you're designing on social media or if you're reading blogs or articles and you can't figure out if someone was actually writing this as them or if they're using AI to write it, this is a cool tool to use. Are you seeing Longshot? Is that the next app that you're seeing on your screen? Yes. Wonderful. Look at that working for me.

19:18So I had Perplexity create something else. I had it create a little paragraph about discovering the lost city of Atlantis. This next tool I'm going to show you is called Longshot AI. What I like about Longshot is it tells you if it's not true or true, but it also gives you citations so you can go do the research yourself. Especially with all these AI tools, I don't care what exists and how accurate it is. You should always be doing your own research, in my opinion. So here is the little paragraph. There's multiple claims in here that are not true. And if you can see on the right here, I can select all of them and actually click fact check claims and it will go through and verify the information.

19:54And what it's going to do is it's actually going to tell me where I can find every single claim rebuttal. So it's going to pop up and tell me if it's true or not. And then where I can go do my own research to back what it is telling me as soon as it's done loading. That's amazing. Again, that's one of the biggest problems, right? And like a few people said, it's going to become a bigger and bigger problem as more and more AI generated content floods the web. And so being able to detect what's true and what's not as you're sharing it is going to be a really big part of this. Yeah, absolutely. And I might want to pivot because it's still thinking and I'll come back to it.

20:30Yeah, you can come back to it. Perspective API is that third tool I wanted to talk to you about because it also starts with you as a user putting out content. Maybe it's going to be counterintuitive to talk about this, but you might not know that the content you put out is stirring up information. I like using perspective API because it actually tells you, you can put in the text that you want to say on social media, out loud, whatever. And it will tell you whether or not your comment is deemed toxic or if it could be controversial. And basically if it will start a little bit of disruption online.

21:01So you can actually self-moderate and help go towards the solution and not be as part of the problem. I'm going to bump back here to long shot because it did just load. So over here on the right-hand side. Like I said, it tells me the issue. Obviously, the law city of Atlantis does not exist. Here's the opportunity. So here's how I can revise it to make it true. Here's the source and the reasoning. And then it gives me suggested links to do my own research. And it goes through every single claim that was in this blog. So that's the tool I want to show you. A couple more tools before I wrap this up.

21:32Bot Sentinel. If anyone's on Twitter and you have a lot of Twitter bots that follow you, but you don't know that they're Twitter bots, put their Twitter handle in here and it will send you an email telling you whether they are or not. Again, back to the data. And then Robocall is this last app that I want to hit on. It's actually for your phone because a lot of Trojan horses happen through SMS and not phone calls, which is an unpopular opinion, but it is actually true. I'm not giving away anything, but if you follow me on LinkedIn, I have lots of announcements coming this week in the form of courses and books.

22:00And holy crap, that was stressful, but I did it. Thank you so much. Awesome. This was really great. I appreciate what you were sharing. This was drinking from a fire hose while drinking from a fire hose. You made one session into five smaller sessions, which was amazing. Great tools, great content. If you want more from Ashley, episode 31 of the Leveraging AI podcast. And next we have Susan Fru. Susan is a highly successful entrepreneur who is completely transforming her business with AI. She makes it even more amazing because she's in a low-tech business, which is not expected unless you're one of Carl's clients to know how to leverage AI in order to do that.

22:42And she's going to share exactly that, how she used AI to save$150 ,000 in her operating budget just by using AI. Susan, welcome again to Leveraging AI. Hey, sir. I'm so excited to be here. Thank you so much. And I'm going to give my hour-long keynote in four minutes and 26 seconds. So let's go. All right. These are just some of the AIs that I use. I use so many different ones in our business. But really, it starts with this. The theory is always this. You want to automate the ordinary and personalize the extraordinary. So getting back to what we heard earlier about systemizing things, I took an entire week of business every single day for five days.

23:33And I went through and I re-automated or evaluated every single policy, procedure, process in our entire business and built that on a trainial module so that our entire company, which is not a large company, we're not a large company. So$150 ,000 savings in a$2 million company is massive. And our company was in trouble. When I came back to run our plumbing and heating company last June, we were gasping for air. And the only way for us to really see our way out of that was to systemize things and save a lot of money. So we created all of these policies and procedures, actually 2 ,000 of them in total.

24:18And let me just back this up real quick because that's really important. Automate the extraordinary, personalize the extraordinary, and going through all of your SOPs and seeing where things are broken so that you then have the ability to correct them with AIs. One of the things I did is I built some customized AIs for finances. I used all of our blind data and threw it in with key performance indicators to create a budget based on our industry standards. I was also able to replace our bookkeeping service that we were paying$2 ,500 a month for, get it down by using bookie.ai inside of our QuickBooks based off the budget.

25:03And then I was able to hire a much lower cost accountant at the end of the month instead of having someone outside service almost every single day. Another thing that we did is we put an AI answering service in. We were paying$2 ,700 a month for answering service 24 hours a day. And we got that down and we saved$70 ,000 a year on that. We are using Zyra Talk, which we absolutely love. We have been playing around with the Hey Jen live, going back and forth right now. It's just cost prohibitive for us, but it is something that we are researching. And I trained this model on how to speak about water heaters.

25:44The Zyra talk trained on our data. So it went into our database of our 37 ,000 client interactions over the last 14 years and learned all about plumbing and heating. And it is absolutely amazing how it works. I also published a ton of books to use as hooks using a program called Bookie Draft2Digital. We cut our marketing budget by$63 ,000 using all different kinds of AI tools. And then we created a business, another business called Pet Wash Pros. So we build dog washes in people's houses. And I did the entire business, built the whole thing in one day for under$500. and I mainly used HeyGen generated videos.

26:32I made 40 or 50 of them in a couple of hours and we were off to the races. So in total,$157 ,000 and counting of processes, procedures, AIs, and it keeps going. Yeah, I do think the dog wash is shark tank ready. We build them in your garage, your laundry room and your mud room. And because we're plumbers, we hook up the water and the drain and bam, we took our dogs up camping this weekend and they got to use the dog wash. Amazing. Susan, this was incredible. Again, what you're doing in your business is playbook, right? Analyze step-by-step every single task, figure out what's ordinary versus extraordinary, what could be automated, then categorize them and then go one by one and actually make it happen.

27:24And you're a living proof that this actually saves a huge amount of money for businesses. Best thing is to just follow you on LinkedIn. Follow me on LinkedIn. And if anyone would like a consultation about how they might be able to do this in their business, or if you need a speaker for your conference, I am your person. Awesome. Thank you. Again, if you want more, episode 76 of the Leveraging Eye podcast. Next, we have Johanna Stoffer again. Hey, Johanna. Johanna is a product development rock star, and she's a successful entrepreneur on her own, who is an expert on AI product development and custom automation.