Speaker 1: (00:00)
Hi there. I'm deep Dylan. Welcome to your AI injection, the podcast where we discuss state of the art techniques and artificial intelligence with a focus on how these capabilities are used to transform organizations, making them more efficient, impactful, and successful.
Speaker 2: (00:26)
Welcome back to your AI in this week. We'll be discussing how AI is being used in content generation for marketing. We're speaking today with Kate Bradley Turner, Kate's a co-founder and the CEO of lately and AI powered, social media marketing management platform. Kate, get us started by telling us why did you start lately? What's the story behind it?
Speaker 3: (00:49)
So I used to be a rock and roll DJ, and my last gig was broadcasting to 20 million listeners a day for XM satellite radio.
Speaker 2: (00:58)
Speaker 3: (00:59)
So I mean, you know, part of the art of being, and you know, this as a host of being a good, um, DJ or good podcast host is taking the listener on a ride with you and making sure that they trust you to do so, right. Mm-hmm and making it a two-way street versus a one-way street, even though you wield the mic, how do you make it? So I feel like I have a voice on your show. So when you listen to a new song, deep, your brain must instantly access every other song you've ever heard in that moment. So
Speaker 2: (01:31)
You mean just to contextualize the song or like what you're hearing and trying to make sense of it or what, what would
Speaker 3: (01:35)
You, yeah, what's trying to do is it's looking for familiar touchpoint, so it knows where to index that song in, in the library of the memory of your brain.
Speaker 2: (01:43)
Speaker 3: (01:44)
Okay. And in that moment, because it's drawing and all this history comes forth in nostalgia and memory and emotion and all the things that, that are the, the baseline for trust and trust is why we buy. Right. Okay. Okay. So now come with me. Think about your voice, like a song, your voice is a musical note of sorts. All sound has a frequency. Okay. And when you write me a text or a social media post, or an email, I read that text and I hear your voice in my head. So there's a similar idea here. You, as an author, as a writer have to, or, or the onus is on you to trigger familiar touchpoints in the writing and cue nostalgia, memory, emotion, trust, same ideas. Right.
Speaker 2: (02:36)
Speaker 3: (02:37)
Okay. Now the, the other component here is the theater of the mind, which I love, I think it's the most magical thing ever. Um, so the theater of the mind happens in reading and, and listening. So like when you read a novel, you have to imagine what the characters look like. Yeah. And a good novelist allows you to do that gives you the room to make it the two way conversation.
Speaker 2: (03:05)
That's what you mean by it. Like basically making room for your internal mind to fill in the voices, to fill in the landscape, to like color things.
Speaker 3: (03:14)
That's right. And not only making the room, but, but a good host or a good novelist. So cuz they're both parallel here will make you push you towards the right, uh, storyline also.
Speaker 2: (03:26)
Speaker 3: (03:26)
Right. So I'm controlling the situation, but I'm making you feel as though you're, you're part of the conversation that you've got some free will here.
Speaker 2: (03:34)
Yeah. It's like in, I mean, I actually wrote a lot of fiction in a prior life and um, yeah. It's I get it in the writing world. I've just never sat down and honestly thought about it in the podcast world. yeah. But I suppose, um, you know, that a lot of those similar tactics and constructs sort of apply. So take me to lately, like what's the link to lately between, between this kind, the, the DJ, the music background and the bringing the reader or listener alone.
Speaker 3: (04:04)
Let me get to part two and then answer your question. So part two is after I left radio, I started a marketing agency and I, um, worked with my first client, which is a little company, you know, called Walmart.
Speaker 2: (04:17)
Speaker 3: (04:18)
And I built Walmart, a spreadsheet system that got us 130% ROI year over year for three years, taking what we just talked about from radio and putting it into the marketing that I was creating for them.
Speaker 2: (04:32)
Speaker 3: (04:33)
That was coupled with another thing that I learned at the time. This is 2000 7, 8, 9. So a while ago was that the largest retailer in the world had the same problem as the library down the street of the mom and pop that I was working with, which was fear of the blank page number one, number two, how do we take long form content that we're all spending hours doing? Whether it's podcast like this, or, you know, four hours writing a blog or a white paper, and then repurpose it in a way that can really elevate and leverage those four hours spent max really maximize that time.
Speaker 2: (05:12)
I wanna stop you there. Cause that, I feel like this is a question to you as a marketer. Um, and we'll get to the AI stuff soon, uh, to our audience. Um, we'll need to get there quickly, but what's the point of taking a nice thoroughly thought through long form article and then like splattering it all over social media. If not, to just get people there to read the full article or is that really the only point?
Speaker 3: (05:41)
Well, on social media, there's only two points which is click or share, right? Uhhuh that's, there's only two objectives. I mean, so that's the short answer to your question? The longer answer is I only do this for marketing for lately and I have a 98% sales conversion. So that's the proof of the pudding. And I can explain that, but after this show, I'm gonna ask you for the file. The AI picks out the one liners that it knows will get us the most share clicks. And if I trickle them out over time, so there's a mindset change here. You know, if you're thinking of marketing, as in the moment, buts in seats live, nobody cares anymore. People don't listen that way. They don't watch that way either. But if you instead use it as long tail, by the way, which is how the format I worked in radio works is the long tail of, of marketing and listening. Then you get exponential more clicks and shares then you ever would in the old mindset. Right. So that's the why the why is for ROI.
Speaker 2: (06:42)
So it sounds like you're, I guess I was, um, hoping maybe for a more romanticized idea of the, like the value in the actual so cause like social media posts. Yeah. They want to get you somewhere, but sometimes they just have like a value in and of themselves. Like they can just be witty or funny or entertaining or something. Right.
Speaker 3: (07:02)
Speaker 2: (07:03)
Resha. Yeah. Okay. So maybe that's that's what you mean by sh the share part.
Speaker 3: (07:08)
Yeah. So sharing is all about the ego. The only point of sharing content is to that makes it's. It makes you look good. It makes you look smart the same way. Um, when someone recommended a new album in college to you, and then you shared it, you got the credit of being the tastemaker and sharing on social media is the same way I get the credit for your smart, witty thing, right? Yes. So the more I'm able to push out there, the, the more credit I end up getting in the long run
Speaker 2: (07:34)
For our audience sake, we're gonna get into what it means to take machine learning and grab long form content and distill it into small bite size pieces that, you know, have an effective conversion rate. But before we get there, you, you ran your own marketing firm. I think an agency, can you tell us? And it doesn't have to be old school, but like what, what did you traditionally do before lately? And then let's literally dig in on, on, on the AI version.
Speaker 3: (08:01)
Yeah. So I'm gonna explain what lately does and then I'll tell you how I did it. Okay. Cause I think it'll be easy to understand. So lately first, um, you give it access to your social media accounts and we study your analytics and we look back for a year of any social account you give us access to. And as many as you give us access to, and the AI starts to write a, a writing model that starts to learn your voice based on the analytics it's reading. And this happens in 1.8 seconds. And what it's doing is it's studying the messaging in the posts that got you the highest engagement in breaking down that messaging by keyword idea, phrase, sentence structure. So now I have a writing model that I already know is gonna get you incredible engagement, and then it's just a robot. So you have to teach it.
Speaker 3: (08:50)
So now you feed the brain long form content. The more you feed it, the smarter gets. And we give, we give you multiple opportunities to, to get in there and guide it. The AI on its own is cold. A human on its own is slow, but together in our world, you get a one plus one equals 10 situation here, right? That 98% engagement I told you or conversion I told you about for us. So in this case, once I have the writing model, if it's a blog, for example, lately reads the blog, looking for the ideas and the highlights of the, of the quotes and, and send structures that knows are gonna work for you, pulls out those quotes makes them into social posts with a link back to the full version of the blog. If it's a video, it will do the same thing. It'll automatically transcribe the video. And it's looking for the best one liners you. Or I have said writes them into social posts and clips up the video of you or me saying,
Speaker 2: (09:47)
So the presumption here is that the right, like the best one liners, the snappiest one liners are sufficient as a social media post in and of themselves. Is that yes,
Speaker 3: (09:59)
They're, it's meant to ju get you off of the jumping off point. So my job is to start you at third base, 75% of the way there right now, the AI on its own about 40% of the time actually will produce social media posts that are ready to go, or 99% ready to go, where very little editing is needed. And then 20% of the time you're, you're trashing it. And then another 40 you're like, wow, that is such a good idea. Let me rework it a little bit to, to be really clear. It's, it's designed to do a couple of things to help. You know, what words matter. That's really important for forever. Marketing has been working like this. You have great analytics tools and you have great. Um, so social media marketing, you have great social media management tools, but no amount of analytics or management can Polish a turd, right? So if that's what you have to begin with, if you're not understanding what to write about what ideas are actually resonating with your audience, then all of this is pointless. And so we start at stage one, let's find out what's meaningful because meaningful engagement is what gets you, not only the conversion, but the evangelist and that's the business I'm in. Like that's my Uber power in radio is making listeners and fans or customers into evangelists.
Speaker 2: (11:22)
So do you have a young company that maybe doesn't have the audience that they want? They have like a small audience. Do you get enough, uh, signal to even know what topics resonate? And then related to that, how do you separate advertising driven traffic versus organic driven traffic so that you aren't just creating a feedback loop of like mistaken ads being run,
Speaker 3: (11:47)
Right? So the first question is, um, lately has multiple layers of best practices to pull from. So it doesn't really matter how good you are as a marker. So yes, we look at your personal analytics first, but for example, if I was giving you a demo and don't have any access to your analytics, what I'm gonna pull in a demo is gonna blow your mind. And, and I mean, that not lightly. Um, and the reason is, is this, so like you had some fiction writing experience, you mentioned some writing experience. I was a fiction writing major, and I also wrote hundreds of commercials in radio Uhhuh, and I'm good at social. So lately first studies me as its best practice. And I broke down how I write into a series of two dozen rules and I taught them to my entire marketing team. And so when the AI spits us out content and gets us 75% of the way there, my human team takes the rules and gets us the rest of the way there. And then our AI is learning and learning and learning from us. So now it has my brand channels as a second layer of best practice for you because they get me a 98% sales conversion who the hell doesn't want that. Then the third layer is all of you. We have the best practices of all of our customers to pull from as well.
Speaker 1: (13:09)
You're listening to your AI injection, brought to you by zion.com, that's x.com. Check out our website for more content, or if you need help injecting AI into your organization.
Speaker 2: (13:28)
When you describe these rules, are you talking about the machine learning, um, the punchy sentence extractor, if you will, that it's represented in that model, or are you talking about once you've got the sentences, there's some guidelines to help people kind of take it across the finish line, or are you talking about something else?
Speaker 3: (13:48)
Um, all pre actually Uhhuh. So, uh, and I'll break them down. So the first one is the science part, like looking at the, the sentence structures themselves, um, and looking for patterns, right? Cause that's what algorithms do is looking to replicate good patterns and to stop, stop doing bad patterns. And what lately does, is it surfaces word clouds that you can curate around dates? You know, I wanna look at the word clouds in the month of June, and I wanna look at them only on our Twitter channel. And I wanna look at them only in per relation to my Valentine's day campaign versus my Easter campaign, for example. And you can see the words that worked or resonated for that particular amount of time. And you can tell the AI, it'll say to you like, Hey, deep, the ideas about humans are getting a lot of engagement for your customers. Do you want us to look for more content around human specifically or ignore that, count that as a noise word. So that's one way, the second way is the rules. So the rules are pretty, pretty simple. And I'll give you an example of one, one of them would be to avoid using, um, words that undercut your authority like need, uh, just probably think maybe
Speaker 2: (15:07)
Speaker 3: (15:08)
Cause when you do that, it undermines your own trust and people discount what you say. So that's the second thing. And then the third thing you asked is this. So lately has five years of our own proprietary machine language, uh, machine learning and natural language processing that database that we're drawing from. But we also integrate with IBM Watson, Google, Pegasus, um, meaning cloud. And then we were in the closed beta of GT three. So we have quite a lot of, you know, information from being in that previously. Um, so yeah, so there's a lot of things that we're looking at constantly, but you are the most important component, like the human and the AI do not get separated in our world,
Speaker 2: (15:53)
Meaning I'm gonna have editing authority. Yes. And you're really just trying to stimulate what it is I'm writing about.
Speaker 3: (16:00)
Yeah. You, you must actually, because it's, it's the difference between one plus one equals two or one plus one equals 10 that's it's that's the difference.
Speaker 2: (16:11)
And now are you only working with extractive, um, seeds, like you're only extracting actual text as it was written in an article or stated in a thing that subject to transcription or are you playing around with, you mentioned GPT three, some generative models to actually try to put statements together as well.
Speaker 3: (16:30)
Yeah. Um, so we have a closed beta version of the, where, where lately creates content from scratch. You can give it a couple of ideas or guide guidelines you might say, and it will read a blog. For example, it'll pull out the sentences and then rewrite them in your brand voice with mood tone of voice, there's different kind of diff different kinds of nudges. You can give it. So for example, we did a project with Anheiser Bush and Bev, where we studied 10,000 pieces of content. Cause the brain has to learn from something right from them in one of their brand voices. The content came in the form of social media, um, press releases, scripts to radio commercials, anything that this, you know, that drove home, this one brand idea. And then, um, it was able to, in 1.8 seconds, like read a blog and then pull out seven or eight posts using the isms I had learned from the brand. So that is, um, I parked it I did that two years ago and I couldn't afford to integrate it into the product now. Um, cuz I hadn't raised any money and I'm doing the integration as we speak.
Speaker 2: (17:43)
Well, so one question I have is that you must, so you, you do stuff with the MP fours or audio files too, or is it just text?
Speaker 3: (17:51)
Speaker 2: (17:51)
Audio. So, so like taking the audio for example, um, you know, we've, we've done a number of projects, um, you know, where we're passing things through transcription and spoken as you, as you well know, particularly, you know, with your DJ background, but just kind of in general human spoken language is just so fundamentally different from written human language and these transcriptions while, you know, Google and, and Amazon, some of the folks that have access to like really large, um, you know, training repositories, the models are certainly getting better and more accurate. And you do get the words out that the person said, but it just looks horrible when it's typed out messy. Yeah. Yeah. Like you, you can read, you know, you can read many, you know, politicians, transcripts of their speeches and you're like, good Lord. Like how could anyone have thought this was, you know, reasonable to say and you know, particularly the ones that leave the teleprompter. And so how do you deal with those challenges? Because, you know, cause I don't, I, it feels to me like the extractive only approach is just gonna fall apart in that arena. You you're gonna need to get generative.
Speaker 3: (18:59)
Yeah. So good point. So certainly with written content, as you said, it's a lot more instant ratification because generally speaking, we slave over the creation of that content mm-hmm and multiple eyeballs read through it too. Right. Um, with the audio and visual components, we instantly give you a transcription. We work with both Google and Timmy. Uh that's so the transcription is great. It's broken out into voice parts as well, but we give you the opportunity to go ahead and, and delete all the ums and OS and likes and things, the crutches that we all say. So that lately then has a clean transcript to read through before it starts clipping.
Speaker 2: (19:43)
Yeah. I mean the ums and OS and stuff are certainly one of the problems, but the other one is it's not just about the ums and OS there's like multiple cascading problems, like sentences don't quite exist in spoken conversation. Part of it is just this meandering effect. And so I kind of, I can't help, but think that your, like when you go back to your right, your editors, you know, your, your user and they need to kind of clean this up that they have to do a lot of what's gonna happen inevitably in a generative algorithm to try to make it make some sense.
Speaker 3: (20:20)
I think it's thinking about what the objective is. Right? So Lately's job isn't to edit your video yeah. And put it back together for you or to write a blog summary for you. My job is to give you some ideas. So you know exactly what highlights to focus on in your promotion.
Speaker 2: (20:38)
Yeah. And try to link that to some, some analytics in terms of interest in your, from your population, basically.
Speaker 3: (20:46)
Exactly. And now I'm remembering the exact question you asked about paid in organic. Um, the answer is lately you are only giving lately analytics to your organic social posts. You can't give it access to your paid.
Speaker 2: (21:00)
Okay. I want to sort of jump up a level now. Cause I think, and I'm gonna tr I'm gonna try to summarize what, what I think I understand of, of your approach. But I, I feel like there's a, a higher level here that you might be addressing as well. So what I understand is, so you've got this system that given a piece of content, whether it's, you know, audio or, or text, you're able to distill representatives like interesting snippets from it, you're able to sort of rank those based on your understanding of traffic for, uh, like Meyer a particular user's, um, site and content. But there's another question which is like all this stuff around, you know, how often do I post, when do I post things that aren't so much about just pulling out the extraction piece, but you know, there's like posted 6:00 AM, east coast time, you know, if you're in the states of Canada, you know, on Tuesdays, there's like all this other stuff, are you, what are you doing in that arena? Like what do you find works with respect to social media rebroadcasting of longer form content?
Speaker 3: (22:08)
Sure. So lately enterprise platform is a full service social media publishing platform where you can push a button, get 40 social posts from the AI, maybe spend a few minutes, augmenting them, push another button, schedule each one to go out. Let's say once a week, over 40 weeks on Facebook and then push another button that says and optimize them for the time that I should see the most engagement based on the analytics we have.
Speaker 2: (22:37)
How do I know to do once a week versus 10, a day versus a hundred a day versus one every couple weeks? Like how about that? Is that all just up to me as the marketer to go figure out
Speaker 3: (22:48)
We do some tip recommendation with the cadence, but it's not where we put most of our shine.
Speaker 2: (22:55)
Got it. So you're kind of like, you're not trying to tackle that meta space so much. You're just trying to get really good at the content creation stuff.
Speaker 3: (23:04)
In a sense, yeah. Words like what are the words that are the most meaningful to the targets you want to reach and how do you get them? How do you use those words to make them into evangelists?
Speaker 2: (23:16)
Got it. So let's say we fast forward five years, 10 years out, and you know, our machine learning models get better at the extractive side. I'm gonna also assume they get better at the generative side. So they're able to take awkwardly, stated audio statements and get them to be quite high quality. Why is the world better off with this? What good has happened from, from that world and what bad might have happened? And like, are we just all completely oversaturated with really easy to generate content and, and the, a machine learning gets so good that we're just sick of reading anything online, like, like maybe take a step out of your role as somebody running lately, which I appreciate quite a bit like as an entrepreneur and take me back to your poetic fiction roots. Like what did we just do?
Speaker 3: (24:06)
Number one, I don't, I don't think in marketing that humans will ever be replaced by AI, nor should they. And I've said this throughout AI and humans must work together. That's how we have arranged it in our world. Again, the difference is if you want a one plus one equals three equation, or do you want a one plus one equals 10 with humans, you get the 10, it must be part of the collaboration. The second thing is factually speaking, the world will always create more garbage. There will always be more and more content it's cutting through the noise, right? How do you, how are you the one that people listen to that they engage with? Again, this is about making evangelists. No, not nobody, but, but most companies that I've met at any, any industry, I don't care what it is. That's not their mindset. Beep their mindset is make the sale. That's fine, but it's shortsighted. So if your mindset is instead, make a fan, you're gonna win. Not because the fans love you and are doing the work for you and your customers, but because you're listening, you're creating that two way conversation and you have the ability to iterate and learn as the market changes, which does,
Speaker 2: (25:22)
I'm gonna try to read between the lines and take something away from what you said. Um, I think what you're kind of saying is like, look, AI's here. It's gonna get better. Humans have always been here and we need to play better with machine learning. And if we think of AI more as a tool in this context, this content generation context, um, as opposed to any kind of like replacement thing, then we can leverage that to ultimately wind up with better stuff that, um, that is more interesting, more witty, more funny, more engaged, like whatever that will also in that field in essence rise above the garbage that will inevitably just continue to increase in, in, in volume and frequency. Is that a fair distillation of what you said?
Speaker 3: (26:10)
Yes. Cause technology always happens, right? I mean, let's just put AI in that bucket. You can't control it, that you have to move with it. And the pursuit of happiness, the pursuit of betterness as humans is also always here. It's what makes us human. And we learn how to use things to get better technology being one of those things.
Speaker 2: (26:33)
Yeah. I mean, it's a, there's those of us who see machine learning advances as great pattern recognition engines that are getting to be greater, but are fundamentally nowhere near kind of human level intelligence and agree. Then there's another side of the fence that is like, no, you know, the Singularity's almost here and these things are gonna like, you know, whatever. And you know, I'm firmly in the former camp. I mean, yes, we're getting really good with pattern recognition engines, but at the end of the day, like all this text generation stuff are basically just kind of clever parrots that read a lot of stuff.
Speaker 3: (27:12)
Yeah. I mean, GP D three is a great example of that, right? So there's about, you know, 50 or 60 text generation companies that have taken the engine of a car and just painted their own color on it, which is what they all do. So they're, they're not very good and they can't morph on their own because they only have one engine in its G three. So they can only change as it changes to your point. Those companies are making it possible for a lot of people to, to go in and create garbage.
Speaker 2: (27:43)
Although, I mean, to be fair, like I'll jump in on open AI side. It's a real pain in the butt to get token over there. Like they try really hard to figure out what you're up to. They don't just hand out their tokens to like anyone trying to generate all kinds of crazy spams. So like,
Speaker 3: (27:57)
Yeah. I mean, but
Speaker 2: (27:58)
I don't know that they're gonna win that battle to be honest.
Speaker 3: (28:01)
Yeah. I mean, it, it is. I mean, people have been lined up for a while. It's it? It's, what's interesting. The most interesting thing to me is this is the core of what you have gotten to a few point a few times here in this conversation, which is that, um, marketers are lazy. People are lazy, they hate writing. They're terrible at it. They wanna push a button and, and have magic happen. And I I'm so sorry to say that magic. Isn't real. And I'm a big Harry Potter fan. And I reread those books every summer and I want it to be real. I really want it to be real, but it's not real. And when you sit down with QuickBooks to do your account accounting, you've got this awesome software. That's helping you along, but you have a mindset. I I'm sitting down to do work. I'm gonna do some work here. But for some reason with marketing, nobody wants to do any work. They just want it to happen. Cause I don't understand it. Yeah. So this is my challenge because like people come to lately and they're like, well, I just wanna push. I have a button, have it be done with me? And I'm like, oh God, it's not, it doesn't work like that. The, the robot has to learn from something it has to, it's the only way,
Speaker 1: (29:05)
Perhaps you're not sure whether AI can really transform your business. Maybe you don't know what it means to inject AI into your business. Maybe you need some help actually building models, check us firstname.lastname@example.org. That's X, Y O ix.com. Maybe we can help.
Speaker 2: (29:26)
I wanna go back to something that, uh, that I, I thought you were saying with respect to G P D three and open AI and, and others. Um, so like you've got, um, you know, folks like open AI that have these really kind of broad and capable. Um, but actually in many ways, very narrow platforms like where millions and millions of dollars go into training, even just one incarnation of a model, whether that's a speech transcription or, you know, like a, a generative, uh, you know, language model, like G B T three, talking about like some new startup. Would you have a recommendation for another entrepreneur specifically? Like, is it okay to build on open AI or do you, you really need to build in-house or do you think you need to figure out how to use the data? Like in a hybrid scenario,
Speaker 3: (30:12)
You have to have it Mo majority of yours, because I mean, investors wanna investing you if you're just a copycat because it doesn't make anything. I mean that, that's the question, the sustainable mode questions, what everybody's thinking about. Um, can someone else just build it better, faster than you? That's right. There's no competition if, if they can't. Um, and why do you wanna join a sea of copycats, which is what all the G D three copycats are, they're all the same product, literally the same product with different names, different colors of the car, same engine. I mean, I think more importantly is if you're a young company, you have to think about what your, you know, what your mission is like, what do you want out of this, this effort? You know, what I'm doing is crazy. I mean, for, for a female entrepreneur to raise capital and venture land, I only have a shot at two, 2.7% chance of getting the funding that's available. And I've raised 3.4 million, which I've had to work. My not only my, but like my knees, my shoulders, everything off the part right.
Speaker 2: (31:18)
Speaker 3: (31:19)
Um, so I mean, the question you wanna ask is, is, is it worth it? And we didn't answer this before. So let me just tell you really why I left radio. I was sexually harassed to the point of where I developed a partial permanent disability out of stress. My body reacted out of stress. And so to this day, I can't type at all without extreme pain. I use voice activated software drag, naturally speaking. And I talk to my computer a hundred time
Speaker 3: (31:51)
And uh huh. And when I left radio, I went to another music related company and I was having the same. Nobody believed anything's wrong with me because I look normal, you know? Yeah. And now I'm, I'm a pain in the cause I needed special accommodations. And my, my dad one day kind of had it and he was like, lovingly. He was like, you can't work for other people. And there's no shame in that was what he said to me. And so in that, that day, it was this, this week, actually three things happened, confluence of amazing things. So my dad said this, I was reading a self-help book because I was desperate. I tried every medication possible, all kinds of weird west east medicines now self-help books, which, which I abor. And then the other thing that happened was my, my husband overheard my, my boyfriend at the time he overheard my dad went and got me guy KA GEI. Kawasaki's art of the start.
Speaker 2: (32:49)
Speaker 3: (32:50)
And I read it, I actually read the first chapter or so where guy says, don't make a plan, just get started
Speaker 2: (32:57)
Speaker 3: (32:58)
So I tossed the book cause I was like, I don't need this thing obviously. And that's
Speaker 2: (33:02)
What he says at the beginning.
Speaker 3: (33:04)
Yeah. It's like right there. You're like, OK. And cause his, the point is there's no map for what you, what anybody does in the startup because it hasn't done before. The point of his book is to inspire you to get off your and get going. Um, and the, the last thing that happened in this confluence week was a couple of customers who were my fans at XM, wanted to meet me and hand deliver a product versus mail it, they happened to be angel investors. I did not know this. And at the end of the lunch, they said, we love you. Here's 50,000 bucks. Let's start a company.
Speaker 2: (33:38)
Very cool. Yeah. Thanks for sharing that. I'm um, yeah, it's, uh, it's, it's always amazing and inspiring to hear what creates the fire inside of an entrepreneur to deal with the pain they inevitably deal with. Well, thanks so much for coming on. I feel like we got, um, we just had a, a really interesting conversation that covered a lot of stuff. Is there anything else you want to kind of, for us to, you know, that we didn't address that you feel like maybe we needed to address or,
Speaker 3: (34:07)
Well, I loved what you just ended with. I think of it as the lawlessness of startup life, you know, this, this is all the rules are being made. Now, what I love about what you're doing is, you know, just in this conversation we've had not only are you informing people, I think you're, you do a really good job of, of coding them into learning with the questions you asked, but you're also doing something you might not even realize, but you're lifting me up.
Speaker 2: (34:35)
Oh, yay. this has been really fun. Thanks so much for coming on. That's all for this episode of your AI injection as always. Thank you so much for tuning in you enjoyed this episode. Please check out a similar article bars that really digs in on automated content generation using G PT three that's at zion.com/articles. As always tell your friends about us, give us review and check out our past episodes at podcast, zion.com that's podcast dot X, Y O ni x.com.
Speaker 1: (35:06)
That's all for this episode. I'm deep Dylan, your host saying check back soon for your next AI injection. In the meantime, if you need help injecting AI into your business, reach out to email@example.com. That's X Y O ni x.com. Whether it's text, audio, video, or other business data, we help all kinds of organizations like yours automatically find and operationalize transformative insights.