I LOVE AI Month With Noelle

to warn you that that's gonna happen. I mean, y'all should know me by now, but if you're new, that's how this goes. Um, and so very excited to have you all here. Welcome to our Friday call. This is our chance, our I Love AI community call. Uh, so excited to have so many of you here. Some of you are new, but some of you are from way back. So super excited to have you all. Today I'm gonna focus on a couple of really interesting announcements and economic kind of impact of these announcements. So there's a few of you in here who can speak directly to kind of your perspective, so I'm hoping you'll chime in, but I'm not going to put you on the spot. Aarti from New York City is here, good to see you! We are well represented. Bassett, oh nice! So I can't see all the names at the same time, so when you pop up you might be like, I was here the whole time, but it just shows up whenever it shows up. So if I see your name and I'm excited to see you, I will just yell it out. Anytime. And that's true for all of you. So thank you so much for being here. Um, all right, I, I keep trying to look through it. Oh yes, so we're live. So every week we come together, and the first little chunk of the, uh, session I like to focus on what's happening in AI today. So many people who are not in our community literally just listen to this to hear kind of Noelle's take on the world of AI. We got some good stuff today. Um, it's going to be very tempting not to click these shiny buttons I'm about to share with y'all, but, uh, I clicked them all so I, I can tell you what that looks like, uh, but then also share with you what's happening in the world of the AI Leadership Institute, where I'm going to be, who I'm talking to, so you get a sense of who are the people that are paying for AI workshops. Some of you are entrepreneurs, some of you are solopreneurs, some of you are trying to get in front of your clients. So I just think it's a good lens to see who's paying people to speak to them and who those audience members are. And I think that'll be really, really exciting. Um, I want to keep an eye on the chat, so I will try to do both, keep an eye on the chat as well, um, as, uh, talk to you all. This is an open conversation, but the first part I kind of make like a news snippet so that we can kind of cut it out. Um, but then we open it up. We're going to do breakout rooms, so you'll get about 15 minutes to hang out with each other. No matter what, just use that time to introduce yourself to people you've never met before, However, sometimes, y'all, some of you come all the time, so you'll get in a room with people you do know. Find out what they're doing, how it's going. Maybe be inspirational to them. Be like, you got this, let's go. Everybody, you know, the world is moving towards human connection, and that's what I think makes us different is like, you could go to YouTube maybe and listen to some of the things that I say, but it's nothing like getting in a room with people that are actually practicing this work. And we have everyone from artists, creators, all the way to financial services, healthcare, educators. So I think there's a lot to be learned. And some of my best lessons don't come from me learning from someone in my field. It's like this adjacent, like, wow, they're doing that in healthcare. I could do that in my business. And so I'll share with you. This is also the day I share with you the things that I'm listening to and that I'm watching and that books I'm reading. So you'll get that at the end as well. Super excited for the challenge. Ali's in the house. What's up? So, so cool. So cool. Okay, awesome. So let's see. Oh yes, the news. So I wrote down a couple. One, many of you know I signed my second book deal with Wiley. Yay! Which is the second edition. So this is Scaling Responsible AI. That book launched last year and I am now working on its sequel, which is Scaling Agentic Systems. So it's literally the evolution of this. This is like, you know, executives, business owners, They're trying to figure out how to use AI but not hurt their business or their customers. The next one's like, wow, what, what happens when you actually start automating this across business systems and adding autonomy, meaning these systems start behaving all by themselves? And so a lot of that comes down to the discipline and rigor that you have in your business before you implement AI. And some of you can probably guess where I'm going. Most people have very little discipline and rigor in their business right now. Like, Even the big companies, they don't have the discipline. And so when they start automating things, all heck break loose, right? You saw, uh, dealerships selling cars for a dollar, airlines refunding willy-nilly, like all sorts of nonsense is happening, but it's all because they are so excited about baby tiger mode that they are haphazardly moving way too fast into agentic autonomy. Uh, and that's what the new book is all about. That being said, Microsoft, as if they knew, created a new kind of term in the industry probably 6 months ago now, but they just did a huge thing at their Ignite conference at the end of last year on this concept of frontier firms. And so when I drafted this outline last summer and submitted it to Wiley, it also included this concept of frontier models, frontier firms, frontier— like this term frontier? Well, fun fact, yesterday OpenAI decides to launch, right, a brand new platform for enterprises. And guess what they called it? Frontier. I was like, it's as if you know me. So it's really cool. So I'm excited because my book will— of course, it's talking more about the essence of frontier models in the world of machine learning and what that means. For those of you not familiar with the term, it's those who are kind of pushing the edge, right? Delivering the next best value. So if you're a company, if you wanna be a frontier firm, you're the kind of firm that's always looking to use the best model with the latest capability to deliver the highest value for your customers. If you're a frontier model company, that means you are the one building the models that those frontier firms end up using. So think to yourself, like, of course, OpenAI, easy, easy place to go for frontier models, but other companies like Anthropic, Cohere, maybe lesser known, but one of my favorites, Hugging Face, right? Hugging Face is the land of frontier models. Like anyone— I mean, I, I guess Cohere is pretty cool too because Cohere, they like crafted the art of fine-tuning models, creating custom models for custom solutions. They've been doing this now, I don't know, 5— since 2022. So they've got like some good experience, right? 4 and a half years of experience doing this. But what I like about their story is when they started, nobody— everyone said that they were not gonna be successful, right? Everyone's like, who needs a fine-tuned model when I have a big LLM, right? Everyone thought that LLMs were gonna solve all problems and they just stuck, stuck to their guns and like, no, we really do truly believe in small language models solving it all. So really cool. Oh, awesome. Joe. Oh, cool. Joe. Yes. With Gemma 3. Oh my gosh. So, So Frontier, something to look up. Um, you can actually— well, maybe I'll throw it in the chat. I have it around here somewhere. Uh, uh, uh, let's see if I can find it real quick. Oh, oh, oh my goodness, I am in maximize mode, so let me unmaximize my screen. Oh, actually, I think I was going to share it with y'all. Uh, let's see, let's see, let's see. The OpenAI announcement? They did? I hope so. I just put it in. Otherwise I violated my I put it in the link right there. Oh yeah, awesome. But yeah, I think they did. I hope they did. That would be a disaster if they didn't. My beta testing days might be over if that's the case. But it is, yeah, absolutely. Thank you for doing that. So OpenAI Frontier, I'm going to share my screen really quick so you can kind of get a look of the ecosystem that OpenAI is building. And for those of us who've been following this for a long time, I think it's really fascinating because I've watched each one of these layers get built, right? If you look at this, right, they started first with this concept of like small context window. I, I will tell you the very first thing they did had nothing to do with any of this. All they did was they focused on something that Jeff Bezos focused on with Alexa, which is daily use. How do we create something that people use every single day? Not to create good stuff, not to do things well, right? Codex in 20, you know, 2021, 2022, 2023 even, like, was not great, but that didn't matter. We just wanted to get people in habitual use. It sounds like drugs. Anyway, now they're in habitual use and now we can start adding on these capabilities. But I think it's really useful to look at the world through OpenAI's kind of vision because this is exactly the same vision that I heard talked about. I was at NVIDIA. I got a chance to moderate a panel with Cohere and of course Anthropic, who was a new addition to Microsoft's model catalog. And who else was there? Oh, and Hugging Face. And they were all saying the same thing. But now, you know, it's just interesting to see big companies invest in a big way and launch these capabilities. But what that means is that OpenAI is gonna spend a lot of money telling enterprises that their future is not just building their own agents, but creating an agent stack, which is exactly what Microsoft, customer zero, ended up doing, right? So Microsoft, some of you know, a little news for those of you who don't, Microsoft is releasing to all of their mid-tier enterprise customers and above E-5 is the license type, but everyone in March has an E-5 license, which is most people, um, is going to get Copilot for free. But that new Copilot that they're going to get for free— meaning I say for free, but it's for the price of an E-5 license— before was $320 you had to spend per person. And it was enough for organizations to be like, I don't know about this, I'm not sure, maybe we'll buy. How many of you probably have heard of companies or are a company where you bought We'll buy 10 licenses and see if it takes, right? Yeah. So like, anyway, who's to judge like how Microsoft decide to go to market? But they're about to now unleash Copilot into millions, hundreds of millions of users, and they are doing it because they created a new agentic stack. Looks very similar to this. They call it the IQ stack, right? So first and foremost, it's going to have this concept of WorkIQ, which is basically your M365 data fabric, right? All the stuff. If you've ever used Copilot, and I love Copilot, um, now today as in its newest form, but I've been frustrated with Copilot when I would go and I'm in Outlook and I would go to Copilot in Outlook and say, can you update this email draft? And it'd be like, please copy and paste. And I'm like, But I'm in, in Outlook and, and we're looking at the thing right now. You could just do the thing. Um, anyway, and so for a long time it was this— yes, exactly, completely disjointed, which didn't make any sense. I'm like, Copilot, you were the— you could build the smartest engine. Why? Why? Uh, same with PowerPoint. It'd be like, copy the slide you want, or I can only build new slides, I can't change existing slides. Like, all of this friction. As a matter of fact, there was a term— I shouldn't repeat it because I'm an MVP for Microsoft and I would like to Hopefully it'll stay that way. But there was a term that was coined to kind of identify this kind of sloppiness, right, in how we would utilize this technology. And so, but I'm just here to tell you all of that is changing, right? So Microsoft is launching, has already launched WorkIQ. WorkIQ sits on top of what's called FoundryIQ, which is this agentic understanding, right? You build a bunch of agents and all those agents share knowledge. So now when you go to your, you know, your, uh, receptionist, uh, agent, it knows what your social media has done. It knows what's on your website. Like, there's this inherent knowledge, and all of that sits on top of Fabric IQ, right? So there's all these, like, IQ layers stacked on top of each other that now make Copilot a serious contender in this space, which I think is why— oh, I imagine that OpenAI was hanging around during those announcements at Ignite and was like, how hard would it be for us to come up with an enterprise solution to compete with this, this platform? And so it's really, really cool to see. But what I like about it is this elevation, right? Notice agents are at the core of kind of the, the business logic, if you will. But then these different interfaces going through a web browser, going through enterprise, which is basically Copilot, business applications, right? So you could just— I want you to see these patterns that are happening across the ecosystem. Couple other things though that are happening. One, yes, exactly. Thank you, Rick. That's a great way to think about it and put it. One of the other things now, y'all, some of you know I'm a software engineer by trade. That's how I got into this world is that I was not only good at software engineering, but good at training other software engineers. So that was like my superpower was getting people who didn't know how to code to become coders. And it's kind of cool because I now do that with AI, right? People who didn't know AI, now we're building. That's my, That's my jam. But, uh, GPT launched also yesterday, or maybe the day before yesterday, uh, but just in the last couple days, a new Codex model. Uh, and they did this on the— I mean, I wonder if there's some like undercurrent on the dark web or something that they're all listening to, because they launched the same features back to back to back. Um, but Opus, right, was just also released. Opus 4.6. Um, that's a Claude model, uh, or I should say Anthropic anyway. Um, but both of them are supporting, right, engineering work. There were two use cases last year that were by far the most successful at achieving return on investment or return on time for organizations. There's a bunch that did nothing, but there are two that were the highest level of value generation for companies. The first one, which I'll go quickly through because, well, the second one is related to what I'm talking about, but the first one of course is marketing content generation, right? We are now uncovering that people right now, I would even say last year, cared a little bit about whether or not what they were seeing was AI, but mostly because AI would generate— and Rick, you would know more than anyone— 5 fingers became 6 fingers or 10 fingers, or fingers looked like toes, or eyes were not exactly right, and you had too many teeth and they were in the wrong— like, the content that we were using AI to build was just not great unless you were really skilled at understanding how to prompt a model to do that. Now that's all democratized, right? Now it's very easy to go and remix someone's super sophisticated production. Mute. Um, with production scale, like Coca-Cola advertising kind of quality, can now be reproduced with Build Me. I mean, I just did this. There was a really cute commercial, what, it was a baby cat, and I was like Y'all know where I'm going with this. I was like, uh, just make that a baby tiger. And it was done. I was like, finished. Perfect. Um, so the ability for organizations to get a lot of return on investment in marketing content creation, or at least off the blank page. I mean, I'll just tell you right now, I'm using an agent that's building some of my marketing content, and I did a 28-day Twitter feed on my most common quotes taken from my last 5 or 6 recordings, um, and it generated all— and then I said make it look like me. So, you know, Noelle's colors tend to be red, black, white, right? My, my I Heart AI t-shirt. But like, I just popped it out, and then I could just be like, great, it looks like it's good. Twitter X seems to like it. Let's go ahead and just duplicate that to Instagram. I don't need to be in the world of duplicating content, but I could ask an AI to do it. And so I instantly get hours back that I would've spent regurgitating content and moving it over and putting on Facebook versus Instagram, uh, WhatsApp, Substack. Now I have the ability to do all of those things, but be much more autonomous. And that's what we're gonna focus on, of course, during our challenge. Um, what was the other thing I wanted to talk to you about? Oh yeah, the sell-off. I don't know how many— this is not a financial conversation. I'm not providing financial advice in any way. I do not advise anyone in the spirit of like stocks to buy or whatever. So with that, I say to you, um, there was some interesting stuff happening with the stock market. Uh, one— what's the word— um, one perception or one perspective that has been shared, uh, is that people are anticipating something I get asked on all the time, this AI bubble that we're in and that the bubble is bursting. And what this also relates to, that term, if you see it, also relates to a concept at Gartner, the hype cycle, right? Which is when you hit the top of the hype cycle, which we are right there. On the other side of that is this trough of disillusionment, right? When people are like, that's, you know, you said it was great, but it's not that great. Or you said it would fix my business, it's not fixing my business. You said it was gonna generate revenue, it's actually lost us money. Right? This disillusionment happens and it doesn't matter if it's because they didn't read the book or follow instructions or do what they were told or, you know, get a guide. Like it's, it doesn't matter. It just matters that they all of a sudden are like, AI is not all it's cracked up to be. And that's what a, there's, uh, a little bit of concern because these huge companies are investing massive dollars. I'm talking about Amazon, Nvidia, and Nvidia has its own problems with, like, you know, international trade. But, but there's a lot of these companies missed their numbers for the first time in a long time. And so if you're watching the markets, which I'd encourage you to do at least a little bit, because it foreshadows some of the things we're going to see in the market. But the most important thing it does is it lets you know what your customers— well, my customers— what they're most worried about, right? Like, what happens if OpenAI, you know, doesn't build their data center and now has to throttle all of our requests and that only people paying $200 a month get unthrottled access. Everyone else now gets throttled, right? And for those of you not familiar with the term, right, throttle means to like compress the amount of resources that you get and restrict your access. So if you're using it for free, which I'm gonna just tell you right now, don't do that. But let's say you're paying $20. If you pay $20, you get to, It— do y'all remember these days when OpenAI would have a little banner at the top that says you get 5 requests an hour or 25 requests every 4 hours, right? Imagine that world as the, like, play out of this ecosystem infrastructure war that's happening. Um, we'll see, we'll see what, what's going on. But it's nice, it's interesting to watch, but we're also seeing a direct correlation to markets, which I also think— and I only know that because my portfolio is impacted sometimes in a positive way. I mean, I did— everything's pretty low right now, so if you can buy, you should buy. Um, but that sell-off indicates that I think some, uh, parts of the market are getting pretty quick on the uptake that AI is not gonna— we're not gonna stay at the top of the hype cycle forever. Now, the other part of the trough of disillusionment is that people are hungry for stuff that works. And that's where we as AI Leadership Institute, right, our premium community, that's where we get to be very successful because we get to know what works and go to companies that are hungry for like, does this actually deliver value? All you have to do is keep your promise and they will continually stay with you, pay you, keep you as, you know, as a client. Or you will be, you know, you know what I mean? Like they will be your client anyway. You'll create lifelong relationships with these companies. Because so many people are hype people. They got into the market, they said, I can help anybody with AI, um, and they've never actually scaled anything in production ever. And so in the year that we're going to be doing a lot of operationalizing, I think, uh, yeah, there's going to be a lot of moments where people realize that they've invested in someone or something that can't deliver results. So some of you know this, in my classes I always talk about this, the most important thing you can do right you can do right now today is start getting receipts of the work that you do. And that means going on GitHub and building GitHub projects or building stuff in Lovable and storing it in GitHub so that you have this like repository of things. When you talk to someone, you're like, no, no, I'm not just talking, I'm showing, right? Y'all know this. This is the core ethos of I Love AI and the AI Leadership Institute. We are doers, not talkers. We show, we don't tell. We learn by doing and we demonstrate our work. We show our receipts. That's— it's going to be— I mean, it's, it's already— I think in general, for the last 20 years of my career, my ability to show receipts on what I've done has separated me from everyone because no one else does the work. So it's not hard to be good at this. All you have to do is be willing to do the work and talk about the work that you're doing. Um, oh, awesome! What? Amazing! Yeah, exactly. Oh my gosh, exactly. Oh gosh, crypto. Oh my goodness. Yes. Um, yeah. Oh my goodness. Yes. Okay, cool. Cool. I, I can't get excited about what y'all are talking about. Okay, I have a plan, I have a plan. Um, okay, so Frontier talked about Codex, the new Codex model. Um, who was it recently? I think it was Eve. Eve, who's here right now. Y'all could talk to her about it, but she created something with Codex— or no, not Codex, she created something with, um, Claude Code. Uh, and what I like about these stories is that not all of us are like born and raised engineers. As a matter of fact, Some of us would claim we are the furthest from it. But I don't know if you know this about Claude Code, like the goal of Claude Code is to make infrastructure and engineering irrelevant, not unnecessary. It's still necessary, but it's irrelevant to your ability to build something that you use your words and an AI system is going to, you know, it's kind of like building a car, right? Like it's not that the building of the car doesn't happen anymore. It's just that the humans that used to like You know, use the little drill things on all the bolts. Like a human doesn't do that anymore. And that's kind of where we're moving in the, in the terms of, um, software engineering. I would just encourage you to start writing software even if you don't, like just come up with, worst case, go to ChatGPT and be like, I have a problem. What software could I build? And then just go build it even if no one ever sees it. Just build some muscle memory around the fact that you can use your words to build an app, something that I don't know if you know, some people lost their life savings building an app they wish they could build. 10 years ago, people would spend hundreds of thousands of dollars on an app that never got to market 'cause their dev team was bad. Like, it's really a really sad situation. Like I said, if we're the ones who are doers, not talkers, we have a huge opportunity to help people make their dreams come true with very little effort. Like, if we know how to talk to a machine, We can get a machine to do this work. And then if you ground that in the responsible, transparent, explainable principles that we teach, I mean, we win. So many people, they just don't even know what they don't know. So I'm excited about what's happening. I don't know, you probably don't want to share it. Oh, did you share it? Oh yeah, you did. So Eve shared it. So maybe we'll have Eve maybe next week or the week after, we could have you share kind of what you did. But it was so cool, uh, because what Claude Code, their mission is like, don't let the engineering side keep you from doing the thing that's gonna bring value to someone, right? Like the, the fastest way to value is your idea of what problem you can solve, the words you use to create that solution and get it in front of somebody who's like, oh my gosh, that's amazing, or that's awful. And then you don't spend time on that anymore, right? Jeff Bezos taught me, So many lessons in my career. But one of them was don't build stuff people don't want. At Amazon Alexa, 95% of our roadmap was driven by customer feedback. As a matter of fact, we trolled Reddit for people complaining about Alexa so we could find out what features to build in. 'Cause nobody was talking, no one was saying to, I mean, maybe now y'all are, but nobody was saying, hey, okay, I'm on mute. I hope y'all don't have devices nearby, but you know, Alexa, this is awful. You're doing a terrible job. Please do better. Like, nobody was saying that. They would just stop talking to it. So we had to get very creative in ways of finding out. And today it's faster than ever. So, so Frontier, take a look at that. Frontier firms a huge effort on Microsoft side. Frontier, the product that OpenAI has built. New models, the new Codex model in OpenAI, as well as Opus 4.6. I'm going to say this because maybe a few of you know about the limitations of context windows. We do have a class in I Love AI. So if you're listening to the sound of my voice, we have a class that walks through kind of the basics, right? It's called AI Leadership Essentials Lite. 5, 6 hours of content, but it'll just help you understand some of these terms that many of us have gone through that class or do this professionally. And so you know what I'm talking about. But I encourage you to have some empathy because there are some people who have never heard these words before. So I want to take a minute to talk about them. But when we're talking about like these, these models that we're building, this generation of frontier models and pushing the edge, it feels a whole lot like move fast and break things. I'm just saying, like, it's got that vibe of like frontier. I don't know if you remember. No, none of us remember. But maybe you read about the frontier back in the— I mean, literally, it is embracing the Wild West. Does anyone else see that? Like, I don't think that's a good thing, which is why all of us in this community, we care about how we deliver this. We scale responsibly. We at least have the courage to ask questions, right? I always call them baby tiger questions, right? Like, how big is it going to be? What's it gonna eat? Where's that data coming from? Where's it gonna live? Like just basic questions that'll shift the trajectory of the project that we're either building for our own clients, building for ourselves, or as an employee, someone has to be brave enough to be like, I get that we're on the edge and in the Wild West, but how do we make sure we don't hurt anybody? How do we make sure that this serves these, the same people, you know, not only the people we're building for, but anyone else who Who has this problem? And that is, I think, the challenge today is most organizations— I don't know if you saw, we've, we've had some layoffs recently. Very, very, you know, what is it, 5-digit layoffs at one company. There are other companies that have embraced this and just told, like, when you get hired, you're told if AI can replace what you do, you won't be here very long. Like, that's the mantra. Imagine how toxic is that? Anyway, I, I, I shouldn't have said that. That was actually me providing my very limited perspective. The other option is that you realize I need to tie myself to something in the business that an AI can't do. And there's plenty of human work that, that is qualified under that. So I do think that that is something that is gonna change. And who was it? Julie Sweet, maybe, a couple years ago. She was like, the most important thing you're— and she's the CEO, or was. I can't remember her current status. Status, but she got sick. But anyway, but she had said this as the CEO of a 3/4 of a million people business. And she said, the most important thing we can do is ensure that these humans that we have, that they are building systems and reinventing what work looks like for them right now. Now, this was years ago that she said this. But reinvention became the mode of that company. Not how do I use AI to do what I'm doing now better, but how do I reinvent what I do as a human in my job knowing that I can use AI? And that's what I do think everyone's gonna have to do is kind of reinvent what does customer service look like in a world where AI can do what it does? Or what does, you know, building a business, what does making a billion dollars, like all of these things re— are completely reframe. Framed in the world of AI and agents. Oh good, I knew she was there. Yeah, Accenture. I knew she was there, but I thought she, you know, she got sick and I thought she like took a step back. Anyway, I don't know. Wish her well no matter what. I hope she's there. Having a female CEO was awesome. It turned out not to really change the internal guts of a company, but it was awesome while I had it. But great, amazing. Leader and really cool. She shows up on CNBC a lot. So looking at some of her past content, but especially what she was saying in '23 and '24, which is when we were building a practice, many of us here were— not many, a few of us were part of that practice that got her to realize she needed to completely shift the way the business was being run. And not all leaders are up for that, right? Most leaders are blockbuster. In a world of Netflix, right? Most leaders are like, ah, this AI thing, it's gonna blow over, the bubble's gonna pop, we don't want to be involved in the bubble. And they don't really realize, you know, that it's a Netflix world right now. Not, not today's Netflix, but back when Netflix took over Blockbusters, when, when Netflix ate Blockbuster's lunch. Now Netflix is struggling in its own new battle. They're now the Blockbuster, right, in the world of streaming, which I think there's so many life lessons to be learned as we watch these companies. Okay, what else did I want to show you? I think I was going to show you a couple of things. Oh yeah. So this enterprise feature is only— it's available to enterprise customers, but I just wanted to give you a sense of what is available. And I wanted to show you, I don't know, a couple of things. One, limitations. So this is for that Frontier Model program. Notice here, deploy agents to code and research. Those are not the agents I, Noelle Russell, want to help run my businesses, right? I'm looking for like, I need an EA that understands my calendar and my email. I need a social media manager who's going to make sure I get content out every day. I need, you know, like I need different things. I don't, I don't, I don't need code and research. But good. I'm glad that that's there. And I will tell you, as I mentioned before, marketing was the number one use case last year. The second use case, of course, was engineering, right? And I just was on a call. I think, uh, we— I was on a call maybe with one of you. Anyway, I can't remember who was on it, if you were on it with me or not, but we were on a call and they immediately were like, oh, well, AI is a good conversation to have with the business, to have with the business, but not— our engineering team doesn't want anything to do with it. And I get it because they're like, wait, aren't you going to replace us? Right? So engineering was kind of cut out of the conversation, um, in the beginning. And then right before we did our workshop, they brought one representative engineer to the class. So there was one person who was there, and by the end of the class, they also were like, yeah, it's for the business. We don't need AI to help us build. How many of you have heard that before? We don't need AI to do this. Here's the cool thing. It doesn't even have to be AI. I started teaching software engineering back in Y2K timeframe, 1998, 2001, 2002. I'm teaching Java to IBM COBOL programmers, and I will tell you, when I told them, hey, there's this thing called an integrated development environment, like VS Code for those of you who know what that is, I'm like, look, we're going to do all of our code is going to happen in a development environment, not in Notepad, Notepad++, not in a text editor, but in an in an IDE. And they told me what people tell me now about AI and software engineering. They were like, no one's ever, ever going to be writing production-scale code in an IDE, right? And now there are— what'd they say? 23 million developers use VS Code. And that's just one IDE. There are many of them. So it turns out, in the moment of change, we often underestimate the impact of the thing that we're being asked to do. So I encourage you, even if like me, I don't use code in my business, but I definitely build stuff and publish it to GitHub and use GitHub Copilot to like fix it and just for fun, just to keep my skills sharp on what these, uh, features are because my clients ultimately are using these. Um, the other thing is that there are plenty of use cases developers don't wanna do and have never really liked doing, like Well, y'all can guess. Oh my gosh, Eclipse. What? We're like true friends. I loved Eclipse. Yes, Eclipse, Visual Age Java. What? Oh, good times, good times. But those tools, or the coding piece, if you think about it, most engineers do not want to document their code. I do, but that's what makes me special and unicorn-like. I like people to understand the work I'm doing, but most engineers can't be bothered. Stopping and writing an explanation of why they did what they did, mostly because when we write code in the enterprise, we're not even sure what we did. Like, we don't even know exactly how we did it. And so documenting it, right, is a bit vulnerable for organizations. So that being said, coding is— one, we can help people document their code. That's amazing. The other thing that I really like about it is Testing. People don't really like to do test-driven development. They find it compresses their creativity as an engineer, right? Like, "I just wanna build." But when you build, and I will tell you, any engineer who does test-driven development will never go back because you write your test and then you build to the test. Well, what happens when you start writing the test is you're like, "This is boring and awful and I don't like it." And they never really, by the time they get to build the code, they're completely disengaged, and so the code ends up not being good. Here's why this use case is so important to kind of keep an eye on, like meaning the coding use case, is not because we're gonna have AI write every line of code in our business, though it might be coming. But the, the real opportunity is to say, hey engineer, I want you to focus on building new features, or imagine I want you to focus on deploying code that has no bugs in it. Wouldn't that be something? Right? Like, I, I think someone taught me a long time ago, 20 years ago, that like bugs in production is like job security, right? Like we always, we just keep deploying stuff and if it has a bug, great, I get to stay and fix it. And that has not really changed even though we have AI to help. And so I think there's a huge opportunity for us to do code reviews, code testing, like all these things that are not super fun for an engineer to build, They get to still build code, but there's plenty of work, uh, that can make their code better, more secure, and deploy with less bugs. And so I was talking to, uh, the CISO, Chief Information Security Officer, of Anthropic, and which is the Claude people, and he said the one thing that they realized was the number one use case for AI for them in Claude code was, can we ensure that our code has the least amount of bugs going to production? And he's the security officer. So the reason he said that that was so important was because every bug in production is a vulnerability. Some of you are cyber people, right? Every bug that we put out there, every defect in our code, every hole in our application, every code, you know, every line of code that is never touched or has no, you know, coverage, All of that becomes vulnerability for an attacker to take advantage of. So why not use AI to just close that stuff that the engineers actually didn't see anyway? And that is something that, I mean, Claude's doing it. So I'm, I mean, I should say the Anthropic team is using it in that way. So why not us? Awesome. Yes, yes, yes. Oh yeah, sure. Cursor. There's a couple. Thank you, Sanjay. That's a great, great term. A great tool. Cursor, Replit, there's quite a few. But I also want to encourage you to make sure that you ask these questions, right? So if you're using a tool that is a wrapper of, you know, it basically is wrapping LLM behavior, just make sure that you have transparency, that they are, you know, providing model cards that you can tell. Like, you're using a cute UI, but what's underneath there, right? Don't— in the world of AI that we are in now, there's zero excuse for a lack of transparency. They don't have to show us their code. They don't have to open source it on GitHub, but they do have to provide a model card that shows us how they feature engineered it. You know, what assumptions did they make? And that's the only way you can use it safely is to know how did they build it, who built it, why did they build it that way? What problem were they trying to solve? Because if you're trying to use it for a different problem, you're gonna get a different answer. And that typically is mis— represented as hallucinations when actually it's just an error in the direction the model was asked to go in based on how it was built. Most data scientists are pretty irritated. They're like, no, no, no, we built that model to do this. Oh, I don't remember his last name, but Claude, the AI model, I just found a book from 1923 from Claude, whatever his last name is. If you know it, put it out there. But he's a marketer. And that's why they named Claude Claude to represent that marketer. So its earliest days, '23, '24, were all use this for your copy on your website, in your emails, like use it for marketing cuz we trained it. Thank you Hopkins. That's it, Kev. What? Thank you. Claude Hopkins. Like, isn't that useful to know? Like, oh, you built this for direct response marketers. I'm using, I'm asking you questions that are totally different now. Granted, there are more models now. In, in Anthropic. But back then no one knew, like people would say, use Claude, it's really good at this. But no one knew why. Oh, 'cause of the training data. That is, that is always a possi— like the answer is always there. That's why I say one of the first questions you should always ask is, where did that come from? Where did that data come from? How did it get that answer? Who trained you, right? Who trained you, little AI? Where did you come from? Because that Yes. Con— thank you. Lee. Lee, is it? I hope so. Or Leah. Um, but yes, maybe it's Leah. Leah, thank you very much. Um, but yeah, exactly. Context is everything, which leads me to the last point. The new model has a, I don't remember the exact number, but it's over a million tokens in the context window. That changes some things, by the way. The only reason why we couldn't write a book with, um, an AI model in the past is because the context window was too short. So it would continuously— not that I've ever tried, maybe I did, but it would continuously repeat stuff in like chapter 7. It wouldn't remember what it talked about in chapter 1, so it would start repeating things as if it had never said it before. And so now though, with a million, now it could edit a book right now. Yeah, exactly. Thank you, Rick. Yeah. NotebookLM, 25 million. Yeah. These context windows are getting bigger and bigger. And then I'm gonna, I'm gonna have you all jump into some breakout rooms to talk about this. But what is— what does that mean from an infrastructure perspective, right? We know— I was just in Mississippi last week, got my new Blue Deltas, they're on the way, I can't wait to show y'all. Um, but like, we now have a brand new data center going up in Mississippi. There is, uh, a very obviously huge project going out in Texas. You can throw out a state that, you know, huge data centers. I take videos because I fly all over the country. I take videos, and I don't know if you know this, but when you're on approach to land, you can see all of these data centers, all of these fulfillment centers. They're just like awful to look at, but it makes me want— like, every city has this part of their city that's just blanketed with data centers. And I'm just wondering Um, so here's the good news. Sam Altman, Northern Virginia, of course, right? HQ2 for Amazon. Uh, they've got a lot going on there, like crazy, crazy. Um, but yeah, it's, it's just, I don't know, like, I don't know, I'm just trying to think, like, how does— how did Skynet happen exactly? You know, like, how did we get to a point where we were, like, in trouble? I'm pretty sure we didn't know it. And, you know, in the books And, uh, in, um, 2001: A Space Odyssey— like, I love these books that were written, you know, like in the '30s, '40s, '50s, uh, the golden age of science fiction. If you haven't read these books, uh, from Asimov, Bradbury, Arthur C. Clarke, they are fascinating reads now because we're lit— it's literally like, wow, that's, that's like us right now. Does— and I keep asking people, I'm like, do you— did you read this? Because you're, you're deciding this for us right now. Um, but it gives me something to talk about, uh, and, uh, but I am a little concerned. Like, so, oh yeah, the reason I mentioned that is that Sam Altman made a joke the other day. He was talking to Y Combinator, is that it? Or YPO? Anyway, wherever he grew up as an entrepreneur, he was talking to those people and he said, almost jokingly, ha ha ha, I mean, the next thing we'll do is we'll just eject, you know, GPUs into space in order to not heat the planet up. Hee hee hee. And of course, I don't know if you saw, SpaceX just deployed GPUs into space. So like, it's happening, people. I don't know, I'm a little bit worried, a little bit worried. But I would love to hear you all, your perspective on it. You know, where does infrastructure fit in this? In a world where we're selling to business who does not, do not understand the technical infrastructure, right? We're selling easy buttons to businesses that are going to launch LLMs. Infrastructure and engineering is irrelevant, right? Doesn't have anything to do with your decision to use it, but it kind of has something to do with our society, our economics. I don't know, I'm a little bit worried. So I'd love your perspective on that today, um, but also just give yourselves a chance to get to know each other. Um, let me see if there's anything else I was going to show you around here. I think that was it. The Frontier. Oh yeah, yeah, that's what I was going to show you. Okay, cool. Um, I will tell you, I always like to end with how my agents are performing. I know it, I know it. I try not to even evoke the name Joe. I don't even wanna, um, oh my gosh, exactly, exactly. Isn't that crazy, Sherry? Like, people are talking about this, but it is not far-fetched to think about. Anyway, something food for thought. Um, but I will tell you how my agents are performing. I am in day 6 every month, uh, since I found this platform. And we're gonna talk about this in the challenge, but every month I, I kind of get just more and more impressed about how my agents are performing. So I'm gonna take you here. These right now, this is my— can you see that? All right, let me see your faces. Hopefully you can. But this is my agents, 107 hours saved this month. However, the reason it's so high is because I like literally just put my AI, Social Media Manager to work. So you can see, um, I highlighted it for you, but I love that this is my new best friend. I talk to my EA every day and I'm like processing my emails, finding out what's important, what's not important, right? In this case, right here, 27 emails drafted. When I start my day, the first thing I do is go into my drafts and I get to start not with a blank page of how do I have to respond and what's the context. I get perfectly aligned for me. It saves me so much time. Um, but 197 posts so far this, this month, like we're February 6th. It's pretty cool. Um, right? We def— everybody needs an AIEA, a blog post, a legal assistant. So I created, I wanted to test this out, so I created a mutual NDA for myself and the members of the C-suite and AI Leadership Institute, and I sent it to them and I'm like, tell me if you, what do you think? Um, I, every time I get an NDA, I throw it into my legal associate. I'm like, what's wrong with this? Is there anything I should be worried about? Does it have any weird language in it? Right? Everyone should have these agents. But as many of you know, the more specifically focused your agent is, the better it is at that task. Also, the easier it is to measure its success. So my— these are all of the metrics I'm using. 4 of 6 right now. The other one is a sales associate, which I might just show it to you cuz it's pretty cool. Um, this is my sales associate. I mean, he's going out and generating leads. The only reason why he's not shown up there yet is because this sequence I just reset. So every time I reset the sequence, his hours reset, but it's a sequence. Like these are emails that I'm drafting and sending to people, um, about our challenge right now. Right. And it's so cool. Like, and then it sends an invite, but some of you know, I'm out of connection requests. So that right here is where it all breaks cuz I can't connect to anybody, but I still like it just in case. Um, every so often people on LinkedIn that I connected to 20 years ago, um, or 15 years ago, uh, whenever I first started, they go away or their account gets closed or whatever. And so every so often I have a few more connections. So it works for like 2 days, but it's pretty cool, right? This is so awesome, right? Sanjay. Oh my gosh, I cannot wait to share with you all what I'm doing. I will also— oh, revert. I don't want to change anything while I'm showing it to you. Let me show you my posts, right? So this is my, uh, post for this week. But granted, I'm hyper-focused. I have a mission. I love AI Month. I have a purpose, getting you all to take the challenge and learn about how to build your own AI, right? Like, none of this happens unless I actually have clarity of thought on my side on what I'm doing and how I'm building, right? So, um, but I do like a hockey stick graph, right? You can see here like February 1st, ooh, all the way up to— it's pretty, pretty amazing. So this is what we're gonna cover in the challenge. How can you replicate what I do if you want to? Like, even if you don't, I think it's a good skillset to have because most people are using AI in a way that's not tangibly valuable to them. So I just think it's a really great, great chance. Now we are going to take— it's a— so we usually go to like 5:30-ish. Now's the time where I kind of say goodbye to our, uh, live learning friends on like YouTube and LinkedIn, and then we head into breakout rooms. You do not have to join. However, I would encourage you to think about— some of you have done this work, some of you haven't— but think about if someone went to your LinkedIn page and they saw your profile, and then they see your picture, and then they go and they read that first sentence. Mine used to be very, I will say, um, a little self— what's the word? Like self— where I say a lot about myself, right? Bestselling author, keynote speaker, TEDx speaker. Like, it was all about me. Turns out, though, it's cool and no one's ever really complained. It doesn't really inspire anyone to like, be like, hey, Noelle, hit me up. Like, I'd love to talk to you. So I changed that, that statement to an empowerment statement. And I encourage you to just use this time that we're about to spend together. Use it to just kind of mess around with this, this ability to introduce yourself in 15 seconds or less with, I'm— oh yeah, exactly. Yeah, exactly. Um, but like, use this time to say, I empower, you know, small business owners to scale their business with AI employees through whatever, right? What do you do? Now you have two options cuz some of you don't know what you wanna focus on or don't wanna claim the thing you're currently doing cuz you're about to transition into something else. You can also switch that to, I, you know, this is something that I'm struggling with and what I'm working on, right? What would draw someone to be like, hey, let's have a conversation. I could help you with that. Here's the good news. When it comes to like cultivating business relationships, both sides are valuable. If you are vulnerable enough to say you're struggling with someone, something, not someone, don't do the people, no people problems. Um, but if you're like, no marital, uh, advice, uh, not, not in this call. There's another community for that. But, but if you are struggling with something like in your business or you just got laid off or you're about to quit, or like if you're vulnerable enough to ask about that and then you build a solution and you decide to go on your own, the person who helped you through those conversations is likely gonna be one of your first customers. That's how I am. As soon as people launch things, I'm like, I'll buy it. I mean, If it's 10 grand, you'd be, you better be a good seller. But, but you know, if you launch something and you have a $29 ebook, like I'm buying it no matter what. Like that's the kind of person I am. Um, but that being said, I think there's just as much value being in a, uh, position of kind of I need help as it is to be, I'm a helper either way. And as you all know, it's a wave. Like sometimes we're on the, I can help you. Sometimes we're on the, I need help and it goes back and forth your entire life. So, so wherever you are today, I hope you'll take an opportunity to introduce yourself, say where you are in this situation, what you need. Um, yeah, exactly. Sell me this pencil. Mr. Belford, good guy now, kind of. I don't know, I just saw him the other day. He's good, good dude. Um, I'm a big fan. I'm a big fan. Oh, nice, Joe. Human-centered AI strategist, founder, empowering leaders to use AI with trust, clarity. I love it, right? Because somebody who's looking for that. I mean, maybe, maybe there's a little like, cuz you have a software, like you're an im— you're an implementer. So I feel like you should be like, and I build stuff for you, uh, cuz nobody else does that. But not, not, not, not that I'm like offering, I'm just saying what a little option. Um, let's see. Um, oh my gosh. Yeah. So Kev, our newest AI leadership member, welcome. I hope you'll stay for the Q&A. Uh, cannot wait to see you all on the other side of this. So I'm gonna open up breakout rooms right now. And you can choose to talk about anything that we mentioned today, but definitely start with an introduction of yourself, who you are, what you do, who you serve, or how you need help. Um, in either case, holy mama, we have a lot of people in here. Okay. Um, if you're not gonna join the group, totally fine. So your options are, uh, stay. And so if you're listening to this and you're not in the Zoom, you're about to miss out on the best part of our call. Uh, but you can join us. You can go to noelle.ai. It's just my first name. Ai, uh, noelle.ai. You do have to apply to become part of a free community, which sounds shady, but I don't like bots. Um, I had somebody today who I think they wrote— what did they write? Stan Store in their, um, question. Like, what do you wanna learn? Stan Store. I was like, okay, so while I have a Stan Store, I don't really teach Stan. It didn't say anything about AI. So I was like, delete. I used to be worried that people would not like me. I now realize like this community is so worth protecting that I'm, I'm okay saying no. So if I reject you, just ping me and be like, no, no, I'm a good one. Or just have a good name, like a real name and a real picture. Um, but our community is 2,200 strong and, uh, and they've all been handpicked. Like I, I read every single person who comes into the organization. So if you're here, you made it. Congratulations. But that means don't like be selling people tchotchkes in their DMs. Like otherwise you're out. You could tell other people tchotchkes though. Feel free to do that, but just not here. Here we're here to learn, teach each other, help each other. All right. So I am creating— oh gosh, I almost created one gigantic breakout room. That would've been a bit of a disaster. So here we go. Let's see. I also do my very best to separate the bots from your conversations. So I will do my best. If a bot is in your room, I do think you can kick them out. Um, let's see here. Do, do, do, do, do. Oh, I found another one. See, they're a little, they're a little slick cuz they have your name at first. So I have to like, oh yeah, see, there's another one. All right, here we go. All right, it's looking good. Looking good. All right, off we go. Meet each other, hang out, find out if there's a reason for y'all to connect again later. Um, and I'll bring you back in about 15 minutes and we'll close out the day. Thank you so much for being here. Talk to you soon. Hit the button. Click, go in. Goodbye. Amazing. All right. So if you're here, uh, hanging out with us online and you haven't joined our community yet, it's never too late. Please feel free to do that. Uh, don't forget noelle.ai. Uh, every week we come together, sometimes most of the time on Fridays, unless I'm traveling or, uh, speaking at an event and it causes me to have to reschedule, but most of the time, uh, but we we also have a community. It's an opportunity for you to share what you're learning and also learn something new and meet new friends that are on the same path as you. We have everyone from 12 years old to 85 years old. We have people who are farmers, researchers, agriculturalists, doctors, uh, chiropractors, legal professionals, family office architects, you name it, um, and you are welcome here. So thanks for joining us this week. I hope you found some enjoyment in the content we shared, and we hope to see you in