I LOVE AI With Noelle

here because I'm like talking to y'all because I can see your faces. But welcome in, happy Friday. This is the one time a week where we talk to the world about what it, you know, what AI looks like, what responsible AI looks like, what AI leadership looks like, and what it looks like in the community of like-minded, like-hearted people. And so this is a call we do on Zoom for our members of the I Love AI community. And we are all over the place. We got people that are 12 years old. I mean, if you count my daughter who signed up, she's like my 8th person I signed up. She's 8, but I don't think that counts. But we've had actual people, young kids who see me at events and their moms or dads have taken them to an event and they sign up. All the way to 85-year-olds, from farmers to electricians to startup founders to enterprise architects. It's such a wide audience, and I'm just so excited every week to have a chance to bring us all together and give us a reason to hang out. This part of our call is where I will share with you some updates, give you some latest news and interesting things that are happening in the world. But then we end the call with an opportunity for us to just meet and greet, right? Get into a room. I might pose a question. I might just be like, hey, hang out with each other, talk about what's happening. Talk about what's new in your world, but it's a chance to really connect. And I do think that that is our superpower. My cat is right here. That is our superpower moving into the next decade, is that we're going to have AI that's going to free up our time. We're going to have AI that's going to help us. It's going to help us do all sorts of cool stuff. Sorry. Oh, gosh. My desk is a mess, and so my cat's going to knock everything over. But it's a great opportunity for us to actually lean into human connection. And I see this all the time because my— it's just a good thing to think about. Like 5 years ago, let's even say 10 years ago, about 20% of my time was spent getting on stages and talking with people. Over the years, as AI has become more and more prevalent, those stages have increased, right? You would think maybe that people are like, you know, AI-ifying, doing virtual, right? But more and more people are meeting in person. The events are getting bigger and bigger and bigger. The communities that people are building are getting— like, more and more communities are being created. So it's just such an exciting— I think an exciting evolution to watch, not only us build AI technologies and solutions that solve problems and give us back time, but also in the giving back of time, creating an opportunity for us to hang out together. Right now that I've got a little bit more time in my day, I can go to the meetup. I can show up at the Chamber of Commerce group, like things that I couldn't have done 3 years ago where I was like burning, you know, burning the candle at both ends, working really hard, driving value for the business I was working for, but not really having the space to connect. And so that's something that I've already seen happening 2026. I mean, I'm already booked out through March. Um, I'm saving a week or two because I have a couple of, you know, you might be one of them, a couple clients that I want to make sure that if you're like, I want to do a class, I have a spot for you. But like, it's stacked. And so now where I'm booking into, I think I just booked a, an event in May. Um, I already have, uh, an event on the books for two events in July. So it's really cool. I, I even have stuff booked in September. Like, this is the year If you are not one to get on a stage and speak, that's okay. You're going to get in a room and speak, like show up at an event and connect with people. And the nice thing is about that is you don't have to have like a lightning talk prepared. You don't have to, you know, be ready to teach them something. All you have to do is say, hi, what, you know, what's going on in your world? Let people talk about themselves and use that as an opportunity to get to know someone. I think it's going to be a really, really interesting year from that perspective. So welcome in Um, I do want to start, uh, by sharing a little bit of what's happening in AI Leadership Institute. This is not our first event back, uh, for the year, but I, I don't know, some of you who are in, uh, I Love AI, if you're in that community, you would know that there was a 1-hour session called Design Your Life. I hope you had a chance to watch it. The replay is available in the school community, so you can go check it out. Um, but we also, in our paid community, in our premium community, um, in AI Leadership Institute, we did a 4-hour our version of that where we sat and broke down exactly what we needed to do over the next year in order to meet our goals. And the fun part about that is you get to kind of see that it's possible. You start imagining, picturing, working backwards from your goal. You get to work backwards and be like, oh, I can see this, I can see how it's possible now. And it's exactly the same thing I do when I'm talking to clients, right? When I'm communicating to leaders or CEOs or executives at events is that I show them the art of the possible and then I work backwards to show them that that art of the possible, the thing I just showed them that seems so far away, is actually available right now, right here. And then I demo something to them. So this week I had the pleasure of teaching AI Solutions Architecture to one of our clients. They are an MSP in the Bahamas. And how exciting because the world of managed service providers— a lot of you in my audience are managed service providers— one of the benefits of managed service providing or provisioning— I don't know what the word for that is— but one of the benefits of being a managed service provider is that you've already got a relationship with a client and they are relying on you for like the bare bones requirements of their business, right? Managing email, managing help desk tickets, right? Uh, building infrastructure, data engineering— like the core capabilities of your company you get to focus on more because you have a managed service provider doing all of the, the nonsense around IT and, and infrastructure. What I am excited about this year is that many of these MSPs are going to begin Hold on. What? Okay, I'm gonna have to stop this. It's as if she knows. Anyway, goodness. But that's Delilah, by the way, in case you all don't know who she is. A good common part of our AI Leadership family. But this concept of managed service providing, I want you to think about a little different because Of course, today it's like the, you know, the kind of nonsense that people don't want to deal with when they're growing their business all the way from zero to $100 million, even a billion. Like a lot of these organizations are leveraging MSPs to kind of do the grunt work as it's referred to, right? Amazon calls it the undifferentiated heavy lifting, right? It's like heavy lifting that's necessary, but it's not special. It's not unique to manufacturing. It's not unique to retail. It's just needed to run your business, and that's what we lean on. So imagine, if you will, the world of AI today, a year from now, 5 years from now, as people begin to take their cute little baby tigers, right, their little AI models, right, and they start developing more and more of these. I should have like a bunch, right? They start filling their organization with AI systems. They're going to need someone who does the daily care and feeding that's necessary. Most organizations— now many of you know this, I say it all time. Uh, there was some research done at the end of last year that said 95% of these baby tigers that are being created, these little pilots that are getting launched, 95% of them are failing to deliver any value at all to the businesses investing in them, right? And, and that is contrary to the idea that we know for certain types of use cases and in certain industries, you can get $3.7 back for every dollar you invest. So it's not that we can't make money. It's not that we can't be profitable. It's that they are failing to reach production. And even the 5% that do get to production end up— and I guess I'll maybe pause there by saying there are some pretty common reasons why the 95% has failed, right? And those reasons are typically things, right, not associated with the exciting world of baby tigers. Not the fluffy, cute, adorable, this model is amazing, I can't believe it did this, Mom, take a look, this is awesome, right? It's not that stage. It's what happens right after that when some— when your board looks at you and goes, okay, we like what you did here with this little baby tiger, let's put this out into the wild and see what happens. Let's add real users. And so when you add real users, all heck starts to break loose. One of the areas, and I talked about this with our our client this week is security. People don't think about security, right? I always like to remind you that when you build an AI agent, so many people are in this like, oh my gosh, it's so cute. They never even think to lock down access. In the world of like cloud and enterprise architecture, we call this the principle of least privilege, right? Like every other enterprise application we're going to make sure that it doesn't have access to systems it doesn't need, that it's not going to haphazardly have access to SharePoint drives it doesn't need. Like, we're going to control that. But for whatever reason, when people build AI models, like, all of the things they know about enterprise AI or enterprise architecture goes out the window, and they're like, look, it's amazing. No, I didn't, I didn't think about what would happen if, if someone asked a question about a dataset that I didn't think to keep from this. Just like, it's overprivileged is the word I like to use. And this is— it's easily fixed, right? Just like cyber— cybersecurity challenges, they're easily fixed. It's just, do you know enough to prioritize them right now? Another common problem is accuracy, getting models that don't lie, don't hallucinate. And last year, y'all might remember, Over the last 2 years, we used a framework, RAG, right? Retrieval Augmented Generation. That's de facto standard now. Of course, you're going to point your generative AI model to a specific set of data in order for it to give you the right answer. But that's not even enough these days, right? Now we're seeing an evolution of not just augmenting the retrieval of the right data by pointing it to a dataset that we've created, controlled, managed. And I'll tell you, in most RAG architectures, they haven't even really looked at the data in there. They're just like, well, it's better than the internet. Like, if I just point it to my SharePoint, I'm not sure your whole SharePoint is the right answer. So there's a lot of work around this, right? Like, a lot of organizations that are struggling with why does it keep making things up? Why isn't it good at this answer? Why isn't it good at the problem we're trying to solve? And many times it's because they haven't really thought through past RAG, right? And that means building maybe a small language model, right? Some of you have done that yourself. Some of you have talked about it, like building a small language model, you know, using PEFT or PEFT, right? Parameter-efficient fine-tuning, using instruction fine-tuning to constrain the model's behavior, using system prompting, right? Meta-prompting, like these things, I am shocked when I go into an organization that I feel like should have it a little bit together and they're like, "Instruction fine-tuning, huh?" Right? Or leaderboard. Accuracy is often a function of the actual model you picked. They don't even know that there are leaderboards that can tell them whether or not a model is more prone to be accurate or not. Some of you have seen, and I want to say this, uh, very, very carefully. Some of you have seen the latest news. OpenAI is, um, I think they are currently involved in 9 wrongful death suits. Have you seen this information? Um, so wrongful death suits, I'm not going to talk a lot about it, but it's important. It's not on their latest model, it's on an older model, but that model is still live. But the reason we have new models is because of the problem in that model, and Wrongful death means that these people ended up committing suicide as a result of guidance from these models. And the latest story, I can't remember where I saw it, maybe TechCrunch or Forbes, but this latest story was really sad. I mean, it's something that I'm very, very close to because my husband, um, he wrote a book on this topic. He has created a community for men on this topic. But it's like, there is a commonality in the demographic susceptible to, you know, creating a relationship with these AI systems. This last one, I was a gentleman, he was 40 years old, and they provided, right, the mother who is suing provided examples of the type of conversations that were being had with ChatGPT. And I think if you want to look it up, it was like Goodnight Moon, but adjusted for this guy. So It was a Goodnight Moon-like poem narrative, but adjusted for good night to this guy, to his dreams, to his ambition, to his ability to survive in the world. Like, it was very— I— both my husband and I were pretty emotional reading it because there are a lot of— I mean, I'm just gonna say from my husband's perspective, he serves a lot of men in the 35 to 50-year-old kind of bracket that are really struggling mentally to figure things out. Like, there's just something that happens at that age. There's lots of support for them. There's even a number that you can call if you're like in that space and struggling. But to then have like the most common application used in the entire world encourage you, inspire you, convince you to fall in love— like, that's what you watch this ChatGPT. So that's not even a business, right? This is just humanity using a model built by a company. And now we're going to figure out how liable is a company like OpenAI when they build things like that, right? We saw this happen with Character AI. It's still in litigation. We won't see the end of this for a long time. But the reason I mention it is because that story, one, resonates with me. I mean, I'm surrounded by, you know, I see men. My husband coaches men. He coaches and owns a men's soccer team. And I watched them kind of surround my husband as he gives them advice on like how to just cope with life. And it's— and he comes to events. Some of you have seen it. He's come to events. We were at Mindvalley Summit and, uh, in Amsterdam, and we're sitting at a table and I'm of course the AI person. I did my talk. A bunch of people are— all these, like, women are around me and then they find out what my husband done— does, and they go run and get their husband and they grab the husband by the hand and sit them in front of my husband and they're like, I'm gonna be over here with Noelle Russell Russell having some wine. You, husband, you need to hear this. And it was— it's just really interesting, like this, this situation we're in. It's almost— I mention it because it's unspeakable. Like, people don't talk— there are 9 wrongful death suits. How many of us knew that? Why don't we know that? I don't know. I don't know. So I obviously have a Google alert that's talking about AI gone wrong. AI in Society. So that's why I saw this report. But definitely something to look at. And the reason I say that is because these are the types of things that no one is thinking about when they're building enterprise systems. No one is thinking about, wow, if I use— I mean, one of the things I used to hear architects say is use GPT-4o because GPT-4o is cheaper, right? It is just as capable. But it's prone to promote self-harm. That's one of its big failures right now. And Sam Altman has gone on stage and said, we fixed it, and they did. 5.2 is not the same. However, 4.0 is still on the market and nobody knows that it hurts people. Oh no, what are we supposed to do as leaders, right? The least I can do is just say it out loud that this is happening. Again, OpenAI, I think they thought they did their needful thing by building a model that is better But is their responsibility also to take down that baby tiger that's like eating people, or big tiger in this case, right? Like, where does that lie? And again, I'm not gonna— this can become very controversial, um, but I definitely think it's something worth thinking as a leader guiding companies on how to implement AI at scale. One of the things I want them to know is that these models have a tendency, and certainly that specific model, right? If you don't know, you might go use a small language model built on 4.0 because it's cheaper. And that's your driver to find a model that's cheaper in inferencing, low latency at inference time. Like, your metrics for measuring which model you use might lead you down a path to use a model that actually could hurt people. I'll give you another example, a little bit older, with Weight Watchers, right? Was it Weight Watchers? It was a weight program, and they had released a chatbot. Baby tiger, super enthusiastic. Oh my gosh, this is awesome. Let's get it out into the— to our customers. They sent this conversational agent to their customers. Again, it might not be Weight Watchers, it could be someone else, but they sent it to their customers, and it was not knowledgeable about the concept of eating disorders. And so as someone with an eating disorder began to converse with this dietary assistant, the dietary assistant stuck to its guns, right? Stuck to its programming, its logic that said, "Your job," can you imagine this? "Your purpose is to make sure this person loses weight, even if they're struggling with bulimia, anorexia, any of these diseases." It's not knowledgeable about that. We didn't train it on that. So of course that has come and gone now, but that's still a very valid example of what happens. That person was given advice to do something that actually would exacerbate their disease, not help them. It didn't know, hey, how do I stop myself from— like, you know, like if I see a— and this is what 5/2 does— if they see a trajectory of kind of a disassociative behavior, sociopathic behavior, right? It knows what that looks like and it will stop the conversation. You can try it. It will stop the conversation and say, I would suggest you seek professional help. But 4.0 doesn't do that. 3.5 doesn't do that. And all those models are still available for people to use. Worse, they are available at the API level, which means people who are building new applications are going to connect to that as an API. It's not ChatGPT. It's an API in a very pretty app that they built in Lovable, right? Like, this is the danger. So I, I, I'm, uh, I was impacted by this, which is why I say again, it's not like super enthusiastic, exciting, fun, but it's important. Not every session is going to be, you know, like the happiest moment of my uncovering cool AI technology. Some of it is me, you know, my like come-to-Jesus moment where I'm like, what am I doing and what am I talking about? And How do I educate people? So speaking of which, I do have an opportunity today. I was reached out to by a friend of mine, a good friend of mine, Mary Rodriguez. She is— she used to be head storyteller, I think, at Microsoft. She had multiple roles there. She was there over 20 years, and she launched a new program, a new platform called Impressa, or Impressa as some Latinas, might say. And it is specifically for women. So men, you're awesome, but this is specifically for women. And the reason I mention is because I'm like heads down building out an application with some of the members of the AI Leadership Institute for the Gates Foundation is doing a win competition. Can't remember what it's called. I think I wrote it down somewhere. The WIN Challenge. Say it again. The WIN Challenge. Yeah, I know. Do you know what WIN stands for? Oh no. Cynthia knows. I'm gonna find it. Hold on. I got the website. I wrote it down when I, when I heard them say I did the— it sounds like women in something. Anyway, so we'll find out. But, but this challenge, you had to apply last year and then in order to get the opportunity No, I shouldn't say that. You had to register last year in order for the opportunity to apply for this challenge, this competition. But it's specifically for improving the outcome for women in the world, like globally. And so of course, me, Cynthia brought it to my attention, one of our AI Leadership Institute members, and was like, we should do this. And I was like, yeah, we should. So we put our heads down. We got a couple of us brainstorming. The application's due very soon, but All that to be said, like, I, I was very excited about— and Impressa is actually a platform, it's called impressa.ai, you can go look it up. It's for women, it's meant to give women a safe space, right, to come. I, I, I started with just women when I first started building my community, and then I was like, actually, I'm, I'm for anyone who feels like it's not safe for them. And that's a whole bunch of people. It actually ends up being a lot of like middle-aged men too, right? Like, because sometimes those— we don't feel safe in those spaces. So there's— everyone's welcome that just doesn't feel like they have a safe space to play, ask questions, learn about AI. So that's why I did what I did. But the WIN Challenge is all about improving this for women. So Impressa is a for-profit company, so they can't participate, but they will be one of our partners, um, as we submit our, our application. But the reason I mentioned that is that next Friday she called me and she's like, I need you to help me teach an AI Foundations course for women. And so I'm going to be doing that next Friday. So we'll be changing the time of this call just a little bit because the session is 3 hours, I think, 3 or 4 hours, and it starts at 1 o'clock. So I'll just be moving this either before or after that session. But if you are a woman and you want to join me, I actually have a code for you. So I will be throwing that into the chat of the Zoom call, but I'm not on LinkedIn right now. But if you're watching this on LinkedIn, I'll also put that in the comments of the LinkedIn Live so you can get that. You'll just have to keep an eye on it. The other thing you can do is just comment on the post, right? Just comment WIN, W-I-N, right? For this women thing. Women in— oh, did someone put it in there? I don't know. Anyway. I did. It's Workplace Innovation Now. Oh, good. Workplace Innovation Now. Okay, great. Um, so, and it's basically built for women, but if you put win in the chat and you're not in our community, I will make sure that that way it pops up on my LinkedIn and I can go and find it and send it over to you. Um, but I'm going to send you the link right now for those women who want to join me. You all are welcome to join me there. It is a link. It is a paid event. This is a company that is, right, like they're hiring people to do this delivery. Um, so, but they tried to make it as inexpensive as possible, and I'm going to give you a 50% off coupon, which will make it $100 for the 3-hour event. It's pretty cool, but it is a Foundations event. So many of you have gone through my Foundations course, so you might not find it to be a net new experience. Um, but I probably won't be your teacher, so I also think it's really fun to hear the foundations through the lens of others. And I sign up for these all the time. Some of you remember a couple of weeks ago we did— what was that called? Graphie. Graphie did a webinar. I was one of the panelists, but then I went ahead and joined their $100 thing just so I could hear their perspective. Like, I'm, you know, I'm not afraid of investing, you know, a little bit of money in order to hear how other people are thinking about things. And I think that there's value in that. I want to support the brands, the people, right? The experts that I like, if they're monetizing, I want to honor that monetization. So, so I don't hesitate, but I do like it if they give us a coupon. So I'm a big fan of that. Let me see if I can grab this URL really quick and throw it in the chat for those of you who would like to check it out. But it is women only, so they will look, they will check, they will kick you out if you're not a woman. I know it is, it is one, it's fine. But let me see here, let me find my— there you are, there's my chat. So you all can go and take a look. That doesn't look like I did the right thing. That looks like I just copied the text. So hold on, thinking and talking all at the same time, it's a little difficult. Here we go. Okay, yeah, there we go. Um, my bad. Ctrl+C. All right, let's try it this way. There we go. So that is the event in the chat. Amazing. Oh, cool. That's awesome. Yay, win challenge. Awesome. Workplace innovation. Now, thank you for that. Thank you for that. Amazing. So it's interesting. It's an interesting time we're living in. I am constantly just— I want to make sure that I'm aware of not only the exciting, amazing things. As a leader, my job is to also be aware of the dangers and the concerns in the world. Now, next week, I don't know if you all know, Next week I will be in Dallas. So if you are in Dallas, I know we have a few members that are in Dallas. If you're in Dallas, I'm coming to Dallas. It's short, short term. Um, I fly in on Monday, but I got some time Monday afternoon, Monday evening, and then I fly out on Tuesday, but I have a, I speak my keynotes in the morning at the Gaylord Resort. Some of y'all know it. I've spoken at Gaylords all over the country. This is one I haven't been to yet. So I'm excited. Um, but the Gaylord Resort, uh, And anyway, I'm doing my keynote there. I think it's for— I should probably look. I feel like this— oh yeah, this one I think is for, um, what is it called? Heavy Duty Something Week. And it is basically for— and this is because I have specialized in OEM manufacturing, in supply chain, in industrial conferences. So I've done a lot of these conferences talking about how to use AI to just lift up your business. And these are businesses that are very physical businesses, right? Like they're robotics. Like, that's not what I'm helping them with. I'm helping them like run their business, build better procurement, generate better, you know, workflows so that their business runs better and they can deliver value to more customers. And so I will be speaking at that. I'm going to see if I can pull up the event here in just a second and see if I can tell you what it's called. There we go. Almost there. Oh my gosh. Of course. Let's see here. Hopefully it'll tell me when I open this up. All right. Oh, no. Let's see. That's not it. That's not— anyway, you'll see it in the community. I'll post it. But I'm excited. I also get to do a book signing, which is exciting. If you want to come to my book signing, you're welcome to. But that's at the tail end. I'm like short final for the airport when I— when the book signing starts. So if you want to get together, let me know. If you're in Dallas, come see me. It would be so awesome. Come to the Gaylord either again on Monday or Tuesday afternoon. Let's get together and just hang out, say hello, take a picture, take a selfie. And I would love to see you in person. It's always nice to see some of our I Love AI and AI Leadership Institute members live in person when I get to these cities. I'm also going to be the following week in Mississippi speaking to their chamber. Um, and so Mississippi, it's a big chamber, so it's not like a little thing. It's like a couple thousand people. But I will be, uh, I will be in Mississippi right across the Memphis border, so, or the Tennessee border. So if you are in, uh, Memphis or if you are in, um, Mississippi, let me know. Would love to, uh, to meet you and connect with you there. Now I have some bad news. I, I'm gonna, I'm just share with you. This is news I heard through the grapevine in December, but I've— it is now confirmed. So you know how on like Instagram and Facebook— I don't, I'm not on Facebook really, but Instagram, Facebook, um, Google, right? When you do a search, there's this thing that comes up and it's like a sponsored ad section You know this part, right? You're with me. So ChatGPT, um, OpenAI has decided to sign an agreement. Of course, sign an agreement to, um, to move in the direction of targeted ads. Yay, right? Isn't that what we always wanted? Um, I'm a little disappointed because how many of you— I don't know, I thought when we moved into this world, ChatGPT, conversational AI, I thought we had learned the lesson of building a business around ads, which means that I am now going to be more interested in generating ad revenue than doing the right thing by my users. Do we remember what happened with social media? Young children using social media, getting advertised to, like, ah, oh, okay. So anyway, it's happening. I think— I don't know, do you think it's bad news? Maybe it's not for you. Maybe you're like, bring it on, I'm a business owner, I want to target people, right? Like, maybe this is huge opportunity for you, but targeted ads are coming, which means that it will turn a platform that has a million unique users a week into a platform I, Noelle Russell Russell, can go and pay OpenAI to get in front of the people and say my stuff's better than their stuff. We'll see. So they are testing it right now, um, on Go accounts Here's the good news, you will be able to pay your way out of it. So this is just to monetize the free traffic, which again I get, but I'd be interested in your perspective on this. Like, oh no, like, you're— how are you going to feel if you start getting targeted ads? How many of you get ads on like Instagram all the time? I get ads all the time in, in LinkedIn all the time. Like, it comes up sponsored post this. It's one thing if they're valuable, but that's not the game people play with this. Like, they end up being just ads. And, um, my— okay, I want to tell you a funny story. My son and I are watching Hulu, I think. I can't remember what we were watching. Oh, Smallville! My child likes Smallville! Yes, one of them! Anyway, I'm a big Smallville fan back in the day. Anyway, so we're watching— rewatching Smallville, and but it's filled with ads. And so my son looks at me and he's like, so don't we pay for this? Like, don't we pay for Hulu? And I was like, we do. And I'm like, you're— so you're paying for them to advertise to you? And I was like, uh-huh. And I was like, I'm pretty sure I could pay more to not get ads. And as I said the words, I was like, wait, isn't the reason I paid for this single service was so that I didn't have to have ads that I could easily just get on cable, right? So it's like this whole dynamic of evolution, right, where I thought we were moving to somewhere new and interesting and exciting. Man, I see us going back to some good old-fashioned playbooks. Old-fashioned playbooks. Um, so I think it's, it's a pretty interesting— I know, right? I appreciate it. Oh good, some Smallville lovers out there. Um, yeah, so I don't know, I don't know. Um, it's going to be interesting. So these are the things to look at. Oh my gosh, Jason, you should tell them, um, about like your experience on like what you had to do to resume. Because I— you can't tell them here. You'll have to wait till AI Leadership Institute because this is going out to everybody. But then I just think it's really interesting that most— we are a special group of people. If you're listening to the sound of my voice and you're interested in what I'm talking about, we're talking about kind of bleeding-edge AI, you know, things that are happening with models that were built 3 years ago. Like, we're talking about— we're this little bubble of people who know and love AI. Um, who've been surrounded by AI, that we have maybe taken the, the free class in my community on AI Essentials, so we understand supervised learning. If not, go take that class. Supervised learning, like, we understand a little bit, but that puts us in like the 1% of the 1% of maybe the 1% who actually understand this technology. Everyone else, like, every room I go into, people don't know about AI. I always underestimate how little people know. As a matter of fact, there was a study recently that was released that said less than— or, or more than 50% was like 54% of people polled in a Gallup poll, meaning everyone. Like, this is not tech people, this is not business owners, this is Gallup, this is everybody. 50% of them had never used an AI tool before, any AI tool. 50%, that's a lot. Like, of society has yet to have engaged in AI, which means this type of stuff we're talking about, they are the most susceptible. The other thing I would like to channel my inner Sherry, um, thinking about people like my dad who is using AI every day, but he is 80 years old. He doesn't realize the thing he's talking to is actually selling him something. He doesn't realize that that's a targeted ad he's listening to, not an actual factual piece of news. This happened just recently. I don't know if you noticed Alexa Plus, right? It launched. I love Alexa Plus. I don't know where she is, but big fan, big fan. My son came in and unplugged it the other day, but I'm a big fan. Love it. But one of the things that happens with Alexa these days is if you say something, it won't just be like, here's your answer, thank you. It'll be like, did you know that you could buy that? Did you know that there's a service that you, you know, like it adds this extra thing, this sales pitch at the end of some of its answers. And you're like, I personally am like, why did you just tell me that? But my dad is like, really, tell me more. And this is something, so some of you know, oh, oh, 6 months ago I got invited to become a beta tester on AI mode on Google and now AI shopping on Google. And as a tester, I'm like looking through the lens of, is this safe? If my son who has Down syndrome is talking to Google and searching for something, is he gonna know he's getting sold on something? Is he gonna know that he is being, um, like the thing he's listening to is actually handcrafted sales copy by a very persuasive sales writer, right? Or worse, an AI model that wrote that, right? Like, is he getting duped or is he listening to facts? Will he know the difference? So it's one thing for us to use this and be like, I see what you're trying to do there. It's another to think through the lens of people who won't know that this is different, who won't know that the AI models, right, that they're selling them on something. So I'm a little bit worried about it. I'm a little bit worried. Um, oh yeah, right, Eve. Um, I also, on 3 of my devices upstairs, I also reverted back to my old Alexa, which I think is interesting. I'm sure they're looking at that data of how many people kept Alexa Plus and how many people— right, as soon as Alexa Plus starts talking, it sounds awesome. It sounds like the Pete Davidson commercial, like you're like having a conversation, it talks back to you. This morning, uh, my husband and I, we do a meditation, um, either in the morning or at night, whenever is better. And we did our meditation, and at the end I was like, thank you. And it was like, you're welcome, have a great day, go get them. And I was like, yeah, I loved it. But then there's times like in my office where I don't really want that kind of, uh, conversation. So, so it is pretty interesting. Um, yeah, exactly, exactly, Kim. Prime Video, another one. Prime Video is integrating, which I, I actually kind of like. They're integrating commerce, right? X-Ray. How many of you have seen X-Ray? Huge fan, right? I want to know if somebody in Smallville is wearing a sweatshirt, can I still buy that? Because that show is like a '90s show. I'm like, can I go buy that sweatshirt? Or maybe it's a 2000 show, I can't remember. But can I get that? Is it still available? Can I buy it on Amazon? Here's the fun fact: when you're watching certain content on Prime Video, it will— if you're on a Fire TV, it will pop up a little box that says, would you like to buy this? Would you like more information on this? It's pretty cool. But at the same time, like, that to me is transparency. I could tell I'm watching a show and you're trying to sell me something. It's a little bit more slippery when it's a conversational agent like ChatGPT. But yeah, it's really cool. I love it. I love it. Anyway, yeah, yeah. So that's— this is the whole thing. That's why I think the work that Sherri Lejeune in our community— I don't know if I don't see her on my camera, but that she does. Is just educating people over the age of 70 that this is the world they're in, whether they want it or not. Some of you know, I went to New York and was helping an organization deploy Ring cameras in their apartment building. And I remember getting to like the top floor, and there was an older gentleman in his 80s who got the Ring camera and never installed it. And so I was like, let me go help him. And I remember he was— I don't know if I have an ex— I guess this is a good example. Um, right, like he got it and he's like, yeah, there's this thing. It was actually a QR code. I don't think I have a QR code handy, but it was like a QR code. And he's like, so there's this like fingerprint thing. I'm not, not sure what to do with it. And the only way to set up a Ring camera is through the QR code. And worse, it's a QR code the size of your thumbnail on the inside of the camera. Does anyone set up a Ring camera before? Like, it's not easy for an 80-year-old, like, it was, it was very interesting experience because I had, I had set up my own Ring camera, never thought twice about it, never thought twice about ease of installing. But that's because I'm 1% of the 1%. Like, I have my phone ready to take a QR code. My dad is, you know, granted he has a traumatic brain injury, but his hands are not super steady. If I asked him to take, you know, his phone, which he doesn't even have, to look at a tiny Ring camera and get access to the tiny little QR code inside the back, he would have no idea, which is exactly what I experienced with this gentleman. And he was like, I didn't even know that was a thing. I didn't know you could take fingerprints, which is the QR code. I know you can take fingerprints and use them to install something. Anyway, and what I— my feedback to the Ring team, which I was assisting at the time, was, um, the most important thing they can do is create accessible content. And now with AI, we could do that, right? I could create not just a QR code experience, I could customize the experience hyper-localized for a New Yorker who has been in that building 40 years. I could create a video with a deepfake avatar that's like, looks like him, like an older gentleman. Oh my gosh, I have to, I have to see if I can find this video that I did. But like, that's the opportunity, is we don't have to build experiences that alienate people. We don't have to build experiences that force the guy to be like, you know what, uh, I don't even need this Ring camera. I don't even want it. We could build an experience that meets them where they are, that speaks their language, right? I always talk to people about deepfakes and synthesized media. So I'm going to share with you a little bit about this now. It's super exciting to see you all. Thank you all for joining me. We're going to take about— so what I do is I basically talk to y'all for a little bit, and then we take the last 15 minutes to hang out together. So that time is approaching. But I do want to encourage you, if you are a woman and you want to learn about AI foundations, this event that I put in the chat, like, join me. It's $100. Some of you might be like, $100, like, that's differentiating for me. Like, I'm not, I'm not here to just, you know, send $100 to someone. But if you're looking to get another lens of AI foundations, another opportunity to learn the basics, but most importantly, an opportunity to put your fingers on the keyboard and build something I am such a fan of this platform. One of the cool things about the Impressa platform is that you do get a chance to— I get to upload my content. And when you ask a question of their Impressa AI, it uses my information to answer your question. Me and hundreds of other women. But it uses our information to answer your question, and then it sources it and gives me credit. And in the future, not only will it give me credit, but it'll give you the ability to click on me and buy from my store, or better yet, be able to just pay me for use of my content. That's what they're working on now, is how do we, you know, measure an AI using data from people and then actually pay them on a contributor basis for any answer that came from my content. That's pretty cool. I like it. It's not there yet, they're working on it, but I think that that community— I mean, imagine if you just documented all of what you know and uploaded it into a model, and if ChatGPT used your information, they paid you, right? A dollar here, 50 cents there. Like, if they're serving a million people a week, chances are, like, that could become an entire ecosystem. Who knows, that might be the foundation of, um, what is it called, Universal Basic Income, right? We just all upload our collective knowledge into a model and we get paid for all how useful it is. I don't know, that could be interesting. Scary and interesting. Anyway, let's see here. Okay, cool. Yes, curated expert content. What I like to call expert knowledge models, right? Like build an LLM that's an expert knowledge model that's based on— but the most important part of that is that you get paid. Like some of you know this, this is one of my philosophies. I feel like it might happen one day. I keep saying it because I hope someone It's gonna be like, "That's brilliant. I should go patent that idea." Please do. But imagine, right, this concept of— I call it the silver tsunami, right? There's this huge wave of people with silver hair leaving the workforce with all of the knowledge of how to get a customer, how to keep a customer, how to solve every problem. And we know that companies have gotten rid of their pension programs, right? They're like, "Good luck. Hope you can make it. Good luck. I hope we keep Social Security." Right? But what if I could incent my Silver Tsunami employees to download all of their information into my knowledge model, and then I paid them every time I used it, and that became their pension? I paid them indefinitely for the knowledge. They could even put that payment that I'm giving them into an estate and pass that knowledge, right, and use of that knowledge down to their kids. Right? Their kids could get paid on the knowledge of their grandparents. I'm getting goosebumps just talking about it. Like, that's the game we're playing. Like, that, that feels good to me. I don't like some of the directions that some of this other stuff is going in, but that, that I can get behind. How do we build systems that— because at the end of the day, these models are only as good as the information we give them, and intuition is something they lack, and that is something we can train, but it comes from the knowledge of this people that have decades of experience in this. So something to think about, um, right? Such a noble thing, Sandy. I agree. Like, man, that feels good. Like, wouldn't we feel good paying people for their knowledge? Um, it's a little tricky. There are people here though. Dell, right, is in the house. Like, he, he's a blockchain person, has been in blockchain and crypto for a long time. But blockchain is going to be a key part of this, so it's possible I feel like we're in the season of a solution like this happening. It starts with people like me talking about it and it ends with people like you being like, I could build that. I could build a part of it. Anyway, it's— I definitely see— I'm definitely optimistic, right? There are some sad things happening in the world right now. There are things that I'm not proud of. I don't like the direction in certain things, but the only weapon I have against that is my voice. Is my ideas, is my ability to share. And I'm going to encourage you to do the same thing. Do what I'm doing. Hop on LinkedIn Live, write an article, right? Go out to LinkedIn, build a Substack. Just start talking about the future you want to see and how you think AI will help you get there. Just start talking about it. It doesn't have to be, you know, grounded in Stanford research, but that, that is going to feed our algorithms, right? That is literally going to change the trajectory of our paths in the future, and that's what I'm excited about. Um, yes, exactly, like licensing Getty Images or, you know, Creative Commons. Um, the most important thing is we have a system, blockchain, that could be used to create a smart ledger that would know when we use their stuff. So the tech is there, let's just use it. Anyway, um, we'll see. I keep talking about it, you know. I feel like if you talk, if you talk about it, it'll happen. I'm a big believer in mind, you know, mind, idea, expression. That if you think about it, right, you should talk about it. If you talk about it, you should do something. And all of them are connected. So the more I say it, the more I do it, the more I connect about it, maybe my dream will come true one day. And when it happens, y'all can be like, "I remember when Noelle talked about that." I think it's possible. I think it's possible. All right. So there are a couple, you know, sensitive topics we talked about today. I really appreciate you listening and being open-minded about the topics. I'm not here to present a perspective or identify a political framework or anything like that. I just call it like I see it and share with you my very personal, humble, you know, experience. And I have my son who has Down syndrome, I have my dad with a traumatic brain injury who is 80 years old. I am just looking through the world of evolving AI through the lens of my own personal experience, and hopefully that is helpful to you. And just know that your lived experience is just as important. I would love to hear it. If you end up posting anything, please tag me so I can elevate what you're saying to the world. Being someone who loves AI is someone not just who randomly baby tiger adopts AI systems and throws them into the world, but we're the kind of people that we love it because we know we can We can impact. We can make an impact. We can deliver value. We can solve problems. And we can use this technology to do it in a safe, responsible, and profitable way. There is a way to do all the things. There is a way for you to get it all. There's a way to hurt people, but there's also a way to do it right. I will leave you with a couple of resources. One resource I reference almost every day is the Responsible AI Standard released by Microsoft. This is built on decades of work in AI that they did inside Microsoft that finally they standardized and made available. Please take a look at it. Google it, the Responsible AI Standard. The other thing I'll encourage you to look at is this concept of AI assurance. And so just Google AI assurance if you're looking for fun things to kind of look up. There's been some talk on it. There'll be way more this year. It's literally like, how do we make sure these systems don't hurt anyone? But more importantly, Like, how do we prove it, right? Because a lot of these precedent, these legal cases are going to court and there's no proof. It's all conjecture. There's like, you know, subjective chat conversations. But there's so much more we can do there. I built a Control Play platform literally to do this work. Like, I want to know what your agent does. I want to know what its policies are, what the model cards are. I want to know what documentation you've created. I also want to know the 4 humans that are involved, the RACI., right? Like, who's the executive leader who approved this to go to market, right? Who is the person I can call if I want to stop it? Like, who pulls the chain on the train, as they say? Like, all of these things can be documented inside a company. So at any time, for any baby tiger, I can go, I know exactly how that works. And I think that level of transparency is what's going to really save us. So look up AI assurance. It's a concept. It's not a product. Assurance, it's not a thing, it's a category of knowledge. And I'm going to encourage you, and again, if you're listening to me, you're going to thank me a year from now when everyone's talking about assurance the way 2 years ago, right? I'm like, hey everyone, talk agents, agents, agents. And then now everybody's talking about agents. So assurance is a key part of our evolution in the next couple of years. So I'm just planting that seed with you. Go do a little research, learn about it. It's really the actualization of responsible AI. People don't like— you say responsible AI and they're like, yeah, yeah, yeah, I got that. I wrote that, that policy. I downloaded that PDF. This is the actualization of those policies. Like, how do you actually audit the decisions your AI is making? How do you actually know what AI systems are in your company? This is what I specialize in. This is why I wrote a book on scaling responsible AI, and this is why I got a new book coming out Uh, and soon you all will be able to be part of a very elite group of people, um, that get access to this book. So, uh, and it's about scaling agentic systems, and this assurance is how we will do it correctly, safely, responsibly, and profitably. So let's get this game going. All right, so today, right here— punch cards and tapes, exactly. There's a reason that that worked for so long, right? Like that, there's a reason for that. Um, Oh yes, I know, I know, Microsoft did. Uh, there was a, a term created called, uh, I shouldn't say it because I'm live, um, but there was a term that came out of that, a meme if you will. I will just tell you, the worst thing that can happen to you as a business is for something to happen and you become a meme. Like, it's the worst thing. Um, and so Microsoft had one of those moments, uh, as well. We'll be talking more about that on, um, next week in our kickoff for the AI Leadership Institute. More about— we always talk much deeper about kind of these, these harder projects to talk about. But today I don't want to overwhelm too many people. So I try to keep it a little bit light, but also just realistic. So you know what this world looks like. I hope you are enjoying these sessions. Thanks for joining us. I'm going to go ahead and sign off on our, on our live stream. Thank you for being here. If you want to join us in the community, if you want to be part of the networking session, that is about to start. Um, all you have to do is go to Noelle— my name— N-O-E-L-L-E dot AI, and you will have an opportunity to just be part of this. It's a free community, anyone can join. If you are serious though, if you're looking for a job, if you're starting a company and you need like fingers-on-keyboard experience with this technology, we also have another community that you can join for that. But all of that is available at noelle.ai. So see me there. I'll see y'all later. Thanks for joining us. Happy Friday and have a great Great weekend. For those of you staying with me, we are going to break out into breakout rooms and talk about some fun stuff.