This podcast episode is sponsored by ChiefAIOfficer.com, offering training and certification through the International Association of Chief AI Officers. Interested in a new career or leveling up your value in the marketplace? ChiefAIOfficer.com can help. Welcome to this week's episode of Using AI at Work. Today we're interviewing Marco Jack, founder of secta.ai. If you're not familiar with Secta, it's probably the category key when it comes to AI-generated headshots and professional images that you can use for your social media and personal branding. This is a fascinating conversation. We covered a lot of territory. He's a great guy and brings a lot of knowledge to the subject of how do I start using AI at work. So with that, let's get started. Welcome to Using AI at Work. I'm your host, Chris Daigle. Each week we'll be learning how today's business owners, entrepreneurs, and ambitious professionals are getting more done with smart use of tomorrow's tech. Let's get started. Okay, Marco, thank you so much. We're actually doing this uh recording from our studio in Bastrop. Marco commuted out here from the other side of Austin. And we've had a chance to kind of start talking a little bit about um where we want this conversation to go today. And uh we have the advantage of the fact that he was involved in AI way before November of 2022 when most of us got involved with ChatGPT's release. So I know we had kind of already covered some of that ground, but if you don't mind, let's kind of go back to that and kind of give us a history of how you got involved in AI. And I'm very curious what what led you to AI in particular, even back then.
SPEAKER_00Um I think pretty much I've always been interested in brand new technology, right? So, like many other people, you see something cool and you want to just give it a go. But I think what maybe set me apart from everyone else is even if it was really bad at the time, I would still be, I would see what the potential could be, right? Because all technology sort of improves over time. Um, so what I'm doing now, I'll just maybe tell you a bit what I'm doing now and then how I sort of started off. Uh so basically Sector Labs is our company, and we've uh since beginning of 2023, um, we've been doing our professional headshots with AI, and it's evolved quite a bit more now into a sort of all-in-one AI kind of headshot portrait studio that you can do anything with. Uh, and like evolving a bit more, I'll maybe tell you like a bit about that. But um, how I got into that actually was in 2022 there were AI image models out there. Um, some people may have seen the whole AI art thing that started in 2022. At the beginning, it was kind of quite poor, the quality of the art was bad. Um, and if you try to make a person with the image model always make a dis distorted view of them. And during 2022, one thing that started, there were a few companies that started up offering these sort of uh AI avatars. Um, did you synthesia and stuff like that?
SPEAKER_02Synthesia and things like that.
SPEAKER_00Well, no, no, AI avatars like see yourself as a kind of superhero or as uh as a Van Gogh painting, yes, right? Yes, yeah, and um on Facebook, everybody was doing them for it. Everybody, yeah, yeah, but that's it. And I think even prior to that, there were some people trying it out, and so I saw the potential in that. I was just this is interesting. I didn't yet know what company to build at the time, but sort of a principle I have is I'll always sort of try it and I'll experiment with it because it's only through doing that that you see where the opportunities come in, as you can't really see them at the beginning, and uh so we actually built like a WhatsApp bot. That's where it actually started. Tell me about that. Well, you could just basically start like start like a chat with like a WhatsApp business account, and it would say upload your pictures. Okay, and we thought that kind of made sense because you have your photos on your phone. Um, and we did this in like seven days. We had the bot up and running, uh, you could choose from different styles. There was like a superhero style and a what's in what's in what's in Christmas style? Because this was in uh December 2022. Uh-huh. And um, yeah, and everyone was kind of like, oh, cool. Yeah, you know, there wasn't anyone actually willing to pay for it. Yeah. Uh and my Were you expecting that it would turn into a business or were you just playing with experimenting with it? I think at the time it was like, well, maybe if we find a way to do this on issue WhatsApp, it will go kind of viral. So really just you know, the unreasonable expectations, but more so just seeing just basically kind of play with it. Can we do it? Yeah, um, and my wife at the time, I did her images at the time, she is still my wife, but at that time, and I did her images and I gave her photos back as kind of Wonder Woman or something, and she was like, Oh, cool, you know, what am I gonna do with this? Uh and one day we were actually having a fight, and she wasn't uh speaking to me, we were just you know having some conflict, right? But then she sent me a text, she said, not that I'm speaking to you, but can you do this for my kind of LinkedIn profile? And at that time, it was really bad at doing any kind of photographic representation of you. Yeah, um, really, really bad at it. But I could see that it was getting better, and that's really the sort of theme here is that all of these things, your first experiences may be bad, but I guarantee you it's getting better quicker than you might think. And and sort of that's we started there. And most people, when they tried it out right at the beginning, were like, well, you can't build a product out of this because it's really bad. But within a couple of weeks, we figured out some things and we figured out some more things, and now, yeah, and now we have regenerated over 22 and a half million AI headshots.
SPEAKER_02That's fantastic.
SPEAKER_00Um, yeah, so that's that story. Um, and I mean I can go into more detail, maybe we go, but pretty much I started with AI in 2014. Um, and at that time I had started a company which was not in AI, but it was with a co-founder, and we were like those two people who are like, well, what are we gonna do? What are we gonna build? Something cool, something that we can, you know, make a lot of money doing. Yeah, yeah, yeah. Uh, and actually we tried to hire an engineer, and uh, I got thousands of what's in job applications. Um, and I was like, Can't we do something to kind of filter all of these applications? And that was like the first time we started, but basically doing stuff in AI. And this is of course prior to the current uh architecture of AI.
SPEAKER_02Um so you discovered the problem in your own business, like you realized, oh my gosh, yeah, there's an opportunity here to screen all of these. Yeah. And then that turned into a business.
SPEAKER_00Yeah, so we we sort of just started building a product to actually fix that problem. Um and then we ended up speaking with a company called uh Ad Corp. Um, Ad Corp is a really big HR company, but they mainly in EMEA. Um and they have Kelly Kelly Quest, they do temporary recruitment. I think Kelly is like a US company as well. And we gave them a demo of this, and they were like, this is great because they were dealing with thousands of applications for every job, and they had thousands of jobs going, and so that's where the first uh acquisition came from was this corporate buyout and um gave us like a really big head. Like, oh wow, we can do this easy.
SPEAKER_01This was easy.
SPEAKER_00Uh yeah, uh a bit of uh a beginner's luck. Yeah, yeah.
SPEAKER_02So where were you doing this?
SPEAKER_00Oh yeah, so uh you can hear from my accent probably. I actually lived in South Africa pretty much all my life. Uh I've been in Austin now since 2022. Okay. But before that, I was in Amsterdam for four years, five years. Okay.
SPEAKER_01Yeah.
SPEAKER_02You came to Austin because of the tech environment or just heard it was a cool spot?
SPEAKER_00I mean, uh my wife and I were always kind of looking for a place we could kind kind of call home. Yeah. Um, and my brother-in-law had actually moved here previously, and he said it's just a great place, the people are great. And he was actually right. I just like we we actually love the states. It's just you just feel more kind of welcome here than you do in any of the other country.
SPEAKER_02Did you know that Austin like was a nascent AI environment when you arrived here in 2022? Um because it is like there's a lot of I mean, you're here, I'm here.
SPEAKER_00Yeah, I did know it was a growing startup capital. Yeah, you know, so everyone knew the Bay Area, right? And there's like a look quite a bit of AI stuff going on there as well now, of course. But I did know it was a growing capital and it felt like a place that we could be at at home and and uh and I could build a company. But this is of course post kind of COVID. So what we've what I've seen is that you can be pretty much anywhere nowadays, right? Everyone's just basically working from home. Yep.
SPEAKER_02So was it this um experience with your wife's request that kind of got you thinking about the professional headshot environment of realistic photorealistic images?
SPEAKER_00I would say she planted the seed of an application that we could charge for because there was an obvious need there, right? Like people would go and book a shoot with the the photographer, yeah. And the first thing I did was just check how much that actually costs. And I was like, well, I can do that for a kind of fraction of that price. And um at the time it wasn't good enough, but I tried, I've got still pictures that of myself that I tried because I've sort of become like the test model just because I can do it. Yeah, and I don't have to ask anyone to make pictures of them, so I just make pictures, and it just no one was optimistic about it then because it just the technology actually wasn't there, um, and so you get like a lot of pushback at that time when you try and do things that just the quality is not yet there with the AI. That quality improved so fast, yeah. And and sort of the first thing I did was I just I didn't even have an app at that point, I just had the idea. So I built an online form through something like Tally and you could pay on that form, so it had like a what's name stripe integration, and I put a couple of demo images on that form and I put it online in a bunch of like community. What was this? This was in January of 2023. Okay, yeah. So just after the kind of WhatsApp bot that we tried in that December, we pivoted onto this, and my co-founder at the time.
SPEAKER_02The WhatsApp bot was more novel. It was like, oh, these are fun and cute, but it wasn't necessarily but that's right, yeah.
SPEAKER_00Okay, yeah, we we tried to charge people for it, but they were like, no thanks. Yeah, because they think like the quality wasn't there or the applic the use of it wasn't there? Application, that's right, yeah, because it was the stylized pictures, yeah. You as a painting, you as a as a kind of superhero. Yeah, so it just didn't really, you know. It was just fun, yeah, it was just fun. Um and yeah, and uh started with the online form. And the the the the the part that I knew there's something here, right? Is when we had so many orders going on that form that I couldn't process the images that people had to upload basically fast enough because they had to manually just dump a whole bunch of files into this form, and I had to crop them and I had to you know change these images in order to be able to uh what's in train an AI model to represent them with that image model, and um, I turned off the form and I said here's like a wait list, and I had people sharing the wait list with each other. Wow, so I was like, Well, that's probably an interesting sign that there is something here, yeah. Yeah, and sort of we had to build the plane on the fly. Yeah, we you know we went from the the form to automating the image kind of cropping to building a basic interface, still having the form. So this all happened within like a month. We had the entire app up and running and everything. So are you a tech background?
SPEAKER_02Do you have a tech background?
SPEAKER_00Yeah, I've uh for like the last 10 years, I've just been building all kinds of products. Okay.
SPEAKER_02So are you guys training like uh training your own model? Are you using somebody else's engine for the generation of these images, or how did you arrive to where we are now with the quality that Secta's putting out?
SPEAKER_00So I think everyone in this space will try to pretend like they have their own. They built their own thing. Yeah, they have their own model. And the reality is that there are what's a frontier models. Yeah. Uh in the image space, in the kind of video space, open AI, right? Uh Anthropic. In our case, a company called Stability AI. Yeah, they built the stable diffusion series of models. These are image models. Yeah. So similar to prompting an LLM, where you're like, uh, I mean, I think now your audience probably knows how to make an image with AI. Yeah. But at the beginning, people are like, How do you make an image with AI? So it's basically text-to-image, you know? And so where we have differentiated ourselves, so I guess that's what you're trying to ask, is like where what are we doing kind of differently, is in everything that takes you to that image. So, for example, we do a lot of work on the uploads, the quality control on your images that you upload. Um, it may not seem obvious to most people, like, oh, I'll just dump a bunch of pictures of myself into here so that the AI can recreate me. Yeah. Um, but it's a hard problem. It's still like a hard problem. It's like which images do you basically fine-tune on? Um and there's a lot of work in that kind of pipeline in the augmentation. So let me okay, let me tell you a what's a funny story here. So when we started off, the one thing we couldn't really figure out is how do we get you in the center of the image? It sounds dumb, right? But like um, we would generate 10 images of you, and six of them, your head would be cut off somewhere. Okay, right. So now at this point, there was no technology to the opposition you in the image. So you would try to prompt it, like a photo of Chris sent it in the frame, right? But that didn't really work that well. So the first tech we built was are you cut off or not? Just uh, is there a face in this picture? Is Chris in this actual picture? And and then we would discard pictures, so you would get 200 images in your gallery. We always believed in giving more than you know, trying. Uh, we haven't really been a company where we've been like, for every 10 images, we're gonna charge you X amount. It's been a bit more like here's a couple of hundred pictures, anyways. But that was one of the things, and and you know, sort of like that's where the technology started is how do we get that end experience? And since then we've built a whole bunch of things, and we have actually fine-tuned our own model on top of the open source models.
SPEAKER_02So I first found out about you guys, I think, in August of 2023. And as a matter of fact, folks, if you're looking at the um any marketing around this podcast, the image that you see that I've been using since then has been a headshot that I got from uh Secta.
SPEAKER_00Nice.
SPEAKER_02But now you guys have, I mean, and I was happy with those that quality, but you guys have continued to improve it. Yeah. You recently had, I guess in the past 30 days or so, some um is is there like an enhancement to the models that you've released recently?
SPEAKER_00Yeah, so there's uh in the open source space, there's been a brand new company called uh Black Forest Labs, and they came out with a brand new image model. If anyone here has used X uh Twitter, right? And uh there's an AI Groc, I think it's called, right? Uh you can generate images on there. That's actually using Flux. Uh Flux Black Forest? That's right. And Black Forest are actually ex employees who left Stability AI.
SPEAKER_02Okay.
SPEAKER_00Uh Stability AI had some their CEO that was kicked off, kicked out of the company, I think. Well, I don't know if that's the like official story. Uh but and they had actually left and they started their own company. And so um, yeah, so that's you know, yeah. I mean you're getting better results from flux. Well, so we we are kind of testing that model, okay, but there's a brand new model coming out from Stability AI as well. They in the kind of final stages of basically safety testing it and doing all kinds of things with it. They've been much more careful, I think, in terms of ensuring that the model doesn't produce NSFW images, that the data set is actually trained on, is actually fair, there's copyright, all of those things. Um, so yeah, we're we're in the process of basically the overhauling the entire product. But it's you know what uh Chris, it's like you think you've built it, yeah, and then something happens, and then you're like, oh, great, I have to rebuild it. Yeah, and so this is exactly what's going on now. Is that maybe I could tell you a quick you know story like about this is when when we started off, right? You had text-to-image, which is what we're talking about. Like, give me a photo of actually Chris wearing a hat with our professional clothing, and so a lot of the work went into prompt engineering there, like what keywords, what what what sorry, what kind of tokens do we incorporate in here? And you know, just to try and get you like the right picture out. But in me there were a whole bunch of startups trying to do things on that, and then this open source tool came out called um control net. And what control net was was just like a way to specify which part of the image you want changed or where we want to put Chris in the image. They had a piece of software called the open pose. So open pose was actually one of the models. So open pose is you can pose the person in the actual picture, right? And so like different stances and body. That's right, that's right. Yeah, and so what you then find is like all the work you've done up to that point is now gone. And so this has played itself out again and again. Uh, we were a bit smart about it, so we built our uh inference pipeline, inference meaning like how we generate the actual pictures in such a way that we can just plug and play whatever the academic and open source like communities are doing, and if we want to make any kind of changes, we can actually do that. But as you add more features to your product, it becomes more difficult to rebuild your product. Um, and yeah, so so we we are we are making a big upgrade now. Um but ultimately our goal is to be like, can we anywhere there's a picture of you like so there's a thing that you maybe haven't seen about our company, but we tested a product where we put you in any kind of clothing, so you could find a photo of yourself on the internet, um you could find sorry, a photo of a guy wearing a t-shirt on HM store or any e-commerce store, yeah, and you could grab that photo and you could put yourself in it, and it and it and it actually works. And so we had people very interested from from the kind of fashion space, but then we turned that down because I don't know in like anything about fashion. Yeah, yeah. This is this is my t-shirt. Yeah, like what do I know? Yeah, how do I who am I to disrupt the fashion space? But that is one thing I think that is going to actually change. Like, I wonder if you'll have to go into a store anymore, if you can just try on any any kind of clothing with AI. Is anybody doing that now? There are startups now that have picked up on this. We were early with this. I already had this idea in uh what's in June of last year. I was basically thinking like about this. Um, and but I think they're mainly doing it with the e-commerce stores, right? So, like doing it as instead of hiring a photographer and doing a photo shoot with a bunch of models wearing your clothing, just generate them. Got it. We were trying to do it more from the bottom up, like give you a kind of virtual kind of fitting room that you can try on any outfit anyway.
SPEAKER_02So one of the things that I'm I'm curious about now, I I have some branding or whatever that we're looking to do with our business, so I can't necessarily I it some images I need to look like I'm in an office, and and Secta has done a fantastic job of doing those sorts of things. But what are some of the applications, and I don't know if you can talk about it in advance, but for this tool, because the the number one user, what are they looking for when they go to Secta right now?
SPEAKER_00Well, I think currently they need a picture that will actually match their expertise or the image that they're trying to portray somewhere that online. Could be print, could be on their profile picture. And so the options are they take a picture on their own or they ask a friend to take it. And it's that's quite a labor, you know, uh a laborious kind of process. Questionable quality output. Yeah, questionable quality, or you go to the photographer. The irony with the authographer, though, is like you don't get that much choice. Some photographers even will be like, if you want to change your outfit, you pay an extra $200, which was like interesting to me. I didn't know that that's how they worked. But it actually makes sense because it's a function of time, right? And uh so they they're coming to us just as a replacement what's in for that. But what I'm seeing a lot more of our power users are people like you, I would say, who have a need for constant content creation to support their message or their brand. Uh and we have like thousands of users who seem to be using the AI photos of themselves to support an online post or a kind of video. Um and they found that the engagement increases because they have their what's a picture of themselves there. And that's something to do, I think, with like the face. If you see a post and there's a what's face in the image or in the video, we sort of have a natural tendency to stop. Whereas if it's just like a photo of a chair or something.
SPEAKER_02Yeah. So the people that are using this, are they finding that okay, this is my image, this one works well, or are they getting better results from because I mean you guys will will generate hundreds of pictures when somebody uses the service. Yeah. Are are individuals using all of those pictures, or are they is it kind of like a photo shoot where you scroll through a hundred of them and go, I like these five?
SPEAKER_00I think it depends. The vast majority I'd say are like, I just want to get a good picture. Okay. I'm tired of my kind of wedding photo that I have pretty much like everywhere, and it's like two years old or five years old, or I need to, I've just started a job somewhere and they need a good picture, or I'm going to a conference. So it's a very point-based, like, just get me a good picture. And then the choice just gives them the options on like how do they want it to actually look? And we have a whole bunch of uh instant editing AI kind of features. I think we're the only ones who have as many as we do. Uh everything from adding glasses to your face, taking off glasses, uh changing the smile, giving like yourself hair if you want. Yeah, I mean like Elvis. Uh some of the kind of hair stuff is actually funny. Yeah. But yeah, so the vast uh majority is just a replacement for a photographer, right? Just go and get that picture to upload to your profile.
SPEAKER_02So this leads us into a topic of disruption. I've paid hundreds of dollars for uh photo shoots with professionals, and it was uncomfortable. I don't feel like I'm not the camera guy. It's you know, so that was unusual for me. Um, the cost associated with with it was always kind of like a necessary evil. Um, you know, doing wardrobe changes and all those types of things was just like I'm not like you, I'm not in the fashion space. So that was all an like an awkward experience, but I knew that I had to do it, and it was an afternoon, and then I had to wait for editing, and you know, out of all those pictures that I spent hundreds of dollars for, I would get two or three that were good, and some of them were just absolute trash. This has got to be stepping on the business of some of those individuals that I would have hired for hundreds of dollars. Are are you do you think that you guys have enough of a footprint in this space to be introducing disruption to that business?
SPEAKER_00Yeah, I mean, you could just see it online keywords, right? Just how people search on the internet. Uh the keywords around AI portraits, AI professional pictures, AI corporate photos, growing like crazy. It's just growing like crazy. While the other keywords, uh photographer, professional photographer, near me, headshot photographer, near me, those are sort of flat or going down. Uh but I think it's still early. Um, one thing I thought, you know, building this product, like, oh, everyone will know about it. Yeah. But it takes a while for people to pick up on it. Um, I think there's still a perception that it's bad. So if you look at how like photographers are talking about it, they'll say it's what's in what's it's basically tacky AI. That's the term I've actually seen them kind of use, or it looks it kind of looks fake. Um, and and that is true, it can look fake. Yeah, and at the beginning, when we started to do it, a lot of it did, but it's just getting better and better every day. And you know, we start when when we first started, we had like 20% of people just like give me back my money, yeah. And I I cared about that because it was a sign of like the product is failing. So I'd be like, we're always gonna give people back their money, just because that's a a way for us to feel the pain. Otherwise, if you don't give but give them back the money, then you sort of distance yourself from the pain. Yeah, it's like why did we give away a thousand dollars that today? What happened there? Okay, something happened, or we actually inadvertently shipped something that actually broke something, and so um, yeah, so that's you know, we sort of saw that, and but it's still there, the industry is still saying that, right? It's basically fake or kind of tacky. Um I I I genuinely don't think you'll be going to a photographer for or for all kinds of things. I think a photography will re remain where it's secondary to the experience.
SPEAKER_02Tell me about that.
SPEAKER_00So, like, say you have you had like a wedding. Yes. Uh the the core thing is the actual wedding experience, the photography is a way to capture that actual memory. Yes, okay. Uh, I don't I I don't want to use AI, I do want us to stand all together and take a picture because it's uh it's a bonding moment. Yeah. Um, and so in those areas, that makes sense, it's definitely gonna stay. However, in the area where you just want the output, I think that that's the key thing with AI, right? Is like what people are not understanding, I think, is that the the ability for AI to just give you the final output is just getting quicker, better. It's on all three cost, time, and as quality. All three are like improving at the same time. And at the beginning, it looks like a toy because the quality is bad. Yeah, so you go, oh wow, cost was good. I could do a bad picture of myself in like a minute, and the quality is really bad, but the cost and time were good. So you go, like, well, maybe it's a toy, but that quality is improving so fast, yeah. And so yeah, they won't they definitely the people who offer a photography service or an image editing service or a content, a production service, audio, video, images, I think all three. If you just want the output, yeah, that's what AI is going to actually give you, and they're gonna have to find something else to do.
SPEAKER_02If you're enjoying this episode and want to learn more about how to start using AI at work, we've made it easy for you. For just one dollar, you can have full access to the Chief AI officer community, which will give you additional training, custom software, daily training calls on AI tools, using AI automations, getting more from your Chat GPT sessions, and the business of being an AI consultant. Simply go to chiefaiofficer.com forward slash insiders to accelerate your AI journey. Now, back to the episode. I like that the way of looking at that because although I would imagine with the proper prompts and the proper headshots uploaded, I could say, Give me some image, like you know, like give me it photos of my wedding that hasn't actually happened. I mean, it's it's getting to that, I would imagine, pretty soon, that you could be able to any historical event, any uh concept that you can conceive, you can actually put your images in there, and there you are signing the Declaration of Independence or Yeah, actually, those are two different things, right?
SPEAKER_00So that the one that you're talking about, which is super interesting, is saying, Look, I have photos from something that took place. Let's say it's a corporate like event or it's like a wedding, yeah. And you say, How do I use AI to clean them up or to give me more options? So that will be also happening. You say, Oh, you know, because the problem with real-world pictures is they are not perfect. Yeah, so you might have a photo where this is such a great photo of all of us, and I want to I wanted to make a print of it, but Chris closed his eyes. Yeah, damn it, Chris. You know, yeah, sorry guys, you know, now that's a really bad photo. So AI, maybe we open up your eyes. Yeah. Right. Uh, we can already do that now somewhat, but I think that's where AI will be used as a tool by the photographers to give customers more, you know, the option. And the the same thing, podcast editor will give a whole bunch of kind of variations that are generated by AI. So there's this topic of variation, I think, that is common across the entire space. Let's say you know, you want to be like, here's an email that I want to send to my boss, you put it into what's an you know, or an investor update, right? You you say, give me a couple of variations of this, and so it gives you more options. Yeah. Uh, and that's definitely what is going on now, I would say, is it improves your efficiency by giving you more options.
SPEAKER_02I like that a lot, actually. So are you able to talk about kind of where you see secta going? Or I don't want you to give away any secret sauce or anything, but I would imagine as early as you were, you've been and back in June of last year, already thinking about things before the industry had caught up with, hey, maybe we can do this. Are there some things that you're expecting to see from Secta or from this technology that uh is already in the works that you can talk about now?
SPEAKER_00Well, the one thing we're actually working on now is how do we take real photos? Exactly what we were sort of talking about about uh you taking pictures out there, maybe they're of yourself or of your family. Uh how do we give you the tools to make the correct edit to that actual picture? So you say my eyes are closed, open up my eyes, or I don't actually like that shirt on me. Uh that's actually one thing we're um experimenting with now. It is a hard problem. Uh the other thing we were playing with was um multiple people uh in the same picture. There's definitely some kind of viral or consumer startup of like photos of you and your friends and are fantastical places, but it's really hard to put two people into the same picture from two different kinds of fine-tuned models. Um and then we actually working on it on the one-shot tech. Well, that's what I kind of call it. And what that is is basically giving you a professional picture without you having to upload uh 20 pictures and instantly without having to basically fine-tune a model.
SPEAKER_02From one image, be able to create a model.
SPEAKER_00One image. Uh, and so you'll see the big companies, they're all trying to crack that nut. And that's and that's something that we knew from the beginning that was gonna come, but it's a hard, hard problem. So if you look at basically Facebook, yeah, Facebook has like a meta Imagine Me Now and how that works is you take like a face scan. Uh it's basically just taking photos of your face, and they generate pictures of you, uh, and they don't fine-tune a model to do that. So that means then they're not injecting your images into the model, and you have a crisp model that we have to load every time. I think part of that is just they don't want to deal with the cost, they don't want to deal with the time, they don't want to deal with like all the customer data. But I think the even the AI headshot companies are gonna be gone from that perspective. Where does that put Secta? Well, we just we have to be the ones to uh to uh to actually to actually build that then. And we have to put ourselves out of business.
SPEAKER_02Yeah, so I know there's a bias to the to the answer of this question that I'm gonna ask, but Secta was the first one that I ran across that was doing this.
SPEAKER_00And we were not the first in the world to do it, yeah.
SPEAKER_02Okay.
SPEAKER_00At least the professional one. Uh prior to us, there were a whole bunch that did the anime character and all that.
SPEAKER_02Yeah. Is is this a niche where competition is starting to arrive? Huge. Okay. Huge amounts of competition.
SPEAKER_00Yeah, I think that's what you'll see with, to be honest, like my opinion on this is quite strong in that like most software businesses are going to be outcompeted. There's going to be a like every piece of software is going to become a the commodity that anyone can actually boot up. It's a bit of a, but like everything from enterprise software to like consumer software will be able to be built like on the fly. So you would say, look, like wow. In the next five years, I predict that the cost of building an app will be $20. Yeah. In an afternoon. Yeah. And you'll be able to prompt it. Build me a Slack clone.
unknownWow.
SPEAKER_00And then you have Slack. So then, and if the service is AI and every company is fundamentally a product and a service, sorry, I'm maybe going off the topic here, but is that then that begs the question like, what are we going to see? And we're going to see a I mean, probably an explosion of apps and companies, and because the the barrier to entry has just gone down. And it's not even that you need capital anymore, right? In the industrialization age, like you would have to buy the machinery in order to achieve the efficiency gains. Here, you don't need to buy that. There'll be an employee, an AI employee that does that job, and you can actually rent them by the hour. And so then that you know that's what we're seeing even in this space, right? However, the differences are, I think, is how much you care. So there's a lot of competitors we've got who just plug their API in somewhere because there's a bunch of companies also who offer the service now as a the API, as the engine, or some parts of it. So let's say they offer the what's name, what's name, the the a fine-tuning part of the service. So you'll your customer uploads 10 images, 10 kind of photos of themselves, you send it to a third-party API, they train the AI model, you get the AI model back, and then you run the images on there. And so that's what we saw pretty much within two months after we kind of started, just hundreds of these companies. Using stability as their engine? Oh, yeah, yeah. Or just the API. It's it's hundreds of these companies. Yeah, there's probably like a hundred, two hundred. Wow. I get now, I get an email every two weeks from one of these companies saying, Hey, we've been kind of following you for a while. Do you want to buy us? Yeah. I'm like, no, thanks. Yeah. We'll just keep doing what we're doing. Yeah. But we, I don't know, we always gave back the money, always cared about it, perhaps more than we should have. Maybe we'd be like, oh, this is just like fun. But I I don't know. I mean, I've always just cared that the person has a good experience. I still do like a support on the ticketing system though every day. Sometimes in the car, which is not good. Yeah. I shouldn't do support while I'm driving. But uh yeah, yeah. So the a lot of competition has come up. I think there will be even more competition in AR for every single every single product. And of course, quite a bit of it is just it's just like a wrapper around chat kind of GPT APIs or other what's in the frontier models.
SPEAKER_02How does Sector stay the category king? Would you say that Sector was the category king for this?
SPEAKER_00I would say that what where where where we failed to become the category king is on the can is on the marketing side. Okay. We have the best product, but your best product doesn't mean that much if people can't kind of find out about it. Yeah. And the main channel for this product was SEO. Okay. Whereas we invested in as your word of mouth. We were like, if this is good, you'll tell all your friends about it. It's still our what's a primary channel. I mean, 90% of our business comes from, hey, my friend told me about it. Great. But it's not a scalable channel, it's not like uh it's hard to scale it. Yeah, you know, and so it's a good channel because it's good experience and you know conversions from very high conversions, but uh SEO was like the main channel. So our story is that you know if you heard that quote, like your first-time founders care about product, second-time founders care about distribution. Now that's been extended to third-time founders care about retention. Yeah, yeah. And what's the fourth time founders don't care about anything?
SPEAKER_02Yeah, they're not so this retention that's interesting because that quote was told to me early on. A friend of mine had been um COO at Automatic at WordPress. Oh, nice. And he he told me that because I was like, man, our product's great, our net promoter score is super high, and all this stuff. And he's like, Okay, that's good. But second time founders focus on distribution, and that that's what led to us sitting here now. It was the podcast. Okay, I need distribution for our messaging and that sort of thing. Yeah, so um totally relate to that, and that's actually a quote that I use regularly for my own team, and now they understand the importance of quality has to be there for product and the support that you're giving to your own product and the refunds and the the still being hands-on with interacting with your clients is a strong, like that's not happening at a lot of companies. They're more interested in uh unit economics than they are in relationships and things like that. But you I mean, you've you've disrupted the the uh CV, the resume environment accidentally when you're for in your first startup, and you've continued to see that pattern. So I would assume that that is something that when you're walking the dog or you know, staring off at the the sunset, that's popping into your mind. How do I make sure that we are uh not late, that we're not without being too early, also. So what are you doing to address that?
SPEAKER_00You have to cannibalize. Okay, you have to be own business. Yeah, you have to be that prepared to cannibalize. Like, so the the thing is that it's very hard to do that. I mean because of bias, like your own by like you love your own baby coverage? Just you're you're you have people that are doing jobs that you like if that means we cannibalize, it means I don't need these actual people. Like, imagine that at a scale of a thousand people. Yeah. So you have all a lot of incentives to keep doing what you're doing, right? I mean a typical story of a disruption. Yeah. But ultimately, you have to understand what the customer cares about, they care about this, and there's a whole set of the market who just cares about getting one good picture, updating their profile picture, putting it under, you know, you know, on like a resume. And so that's where the technology is going. And so we are like embracing that. We've been embracing that for the last eight months, just continuously trying out brand new things. I mean, still really bad. Otherwise, if it was good, then we would have uh done that by now. So, yeah, really just have to be prepared to understand that to stay kind of stay kind of relevant, there's going to be a change in your business, and that means that you cannibalize some parts of your product, you change like the focus on the market. Um and the other thing we're doing in the short term, so that's more a kind of long-term thing. So let's say in the short term, what we're doing is we're saying who are the customers who love the product they come most. Um, you know, just like anything, and that's maybe that third-time founders care about that. Retention. Retention implies kind of frequency, and that implies care or delight. So you're sort of looking for people who enjoy it and figuring out why they enjoy it, and how do I find more of those people? So that's what we're doing in the kind of short term. It's just like more of like a focus on the market where this product is kind of constantly kind of useful to them. Of course, there's still there will always be a package for people who just want to get a couple of hundred pictures, choose their favorite pictures, and uh and then they're done.
SPEAKER_02So this seems like a like obviously if I sold supplements or I sold uh I don't know, financial analysis, there's a uh it's baked in this recency and and frequency of consumption of this, right? But with the image side of things, uh how do you manufacture that uh that demand for frequency of use of the tool?
SPEAKER_00It's hard, yeah. So uh the one way we are thinking about it is like how often do you see a picture of yourself? And that's what I was saying is like you may, if you're out somewhere, you're getting a picture taken, you're taking a picture in an as in an ash group, that's where the inherent kind of frequency is. Yeah, you can't actually manufacture it. I think that's most many startup founders or people starting a business, they sort of have an idea of what they want to build. You know, I want to build an online marketplace, or uh there's something that's kind of hot. I want to build a decentralized. company and sort of that's that doesn't that does the the form of the company is implied by the customer not by what you want to build the amount of money that you can charge is set by like the market that you're going after so there's this sort of four kind of fits thing that some people have talked about which is your product your market your business model and your channel and these are like four planes and sort of if you change any one of these like it doesn't fit with this market anymore and that means that this market is not prepared to pay as much and then your like channels that you have you can't do ads anymore so we I mean we kind of we we had that problem where ads was like if we're charging you basically $50 for your gallery ads don't really kind of work because the cost per acquisition was too high. Yeah I mean generally you should benchmark as I could what's what I found out is like in at the US uh you plan for what's in $200 just as a kind of acquisition you say.
SPEAKER_02I mean of course you can do it for for less but for scalable like okay ads will be our channel do you have any type of feedback loop with the customers so that you like you've obviously got power users people that are enthusiastic about getting new images and using those a lot in their marketing or their personal branding or even their just their social life or whatever yeah but are you do you have a feedback loop where those power users and it may not even be accurate because a power user may not necessarily be the ideal avatar because they're doing stuff that the average individual isn't but is there a a communication environment where they're saying oh it'd be cool if you did this yeah I mean we've I I talk to customers that all the time um I go through sort of uh highs and lows of that where I'll just be like I haven't spoken to a customer for like a week or two or a couple of weeks I'm feeling bad about it and uh and then I just turn on like the basically there's a thing that I can just turn on that when a customer what's in them signs up they'll get an email from me.
SPEAKER_00I mean it is an automated email but it is a personal email from my inbox and they have my phone number on there everything's on there so they can reply to it and if they do I engage them on there and pretty much on there I'll say like hey so what what are we kind of missing? What was your experience like I try not to send it straight away because if I send it within a day then I become the the support person. They're like oh how do I do this and then I'm like oh actually they are getting support at the same time with with the team. Yeah but that's like the one way which I do it and um we're actually using AI for that like as well. For the support or for the for the what's in the feedback. I think that's that's one big thing that's going to change is and that's how I see myself kind of losing a job because I'm fundamentally a product manager or and I see my job as how do you shape product so what are we going to build? How do you sync the people to build that and that's like a hard problem because you have to communicate well what needs to be built what aren't we going to build um and how do we ship it? How do we make sure that we can get it on on time and so one of the things is in order to shape a product you need to know what the customer wants where the kind of market is going. I mean AI I've been just basically feeding a whole bunch of customer feedback whole bunch of online reviews whole bunch of survey open-ended questions and I analyze it like that. So I basically fine-tune my own customer yeah I create a customer the persona that I can speak with and I mean of course I don't take that as a like as a that is the truth yeah but I take it as a guide of what I should ask the real people about. Yeah yeah yeah so I think that's where for example in the market research space as data scales you'll be able to talk to your own customer anytime you want I think all product teams should be kind of doing that like okay let's have a quick chat online a lem like this yeah but it's an AI on the other end with our female customers who are like kind of talk about that a little bit so where are you lots of what's in startup ideas yeah yeah where are you getting now I I would love to be able to do that with my own stuff right yeah so you're pulling um where are you finding the I guess training data the the sentiment from the customers to be able to create this kind of cloud of feedback that you're able to interact with? Well um currently emails for sure emails for sure uh intercom which is our okay live chat system yeah uh we have a whole bunch of surveys that pop up here and there in the app there's a there's a survey you get when you get your first gallery okay and it's sort of like asking you good bad experience and so we then follow up with that with like other questions uh all the transcripts from my calls I transcribe those calls so you're actually having conversations like you're not just emailing back and forth with clients you're getting all I'll do a call and I'll do like an hour call with the customer uh and then I'll transcribe that I mean of course I ask them up front if they aren't comfortable with it I'm quite transparent with what I'm gonna do with it and then I ingest all of that into a what's an prompt session kind of thing and I say here are like the trends here's the other data um you know let's have a chat and you emulate as if that's the the actual customer how how much is that informing the direction of what you guys are doing that effort like I think all things with LLMs with chat kind of GPT and Claude and all of that what it does is it unblocks you big time. Like you may have a hypothesis like I think I should build this feature and in order to even get to the point where you like start building it you would now need to set up interviews with the people and that takes time. Like if I email 100 customers I'll get 20 interviews people have the time for it but it will be in like a week or two and then this one cancels and then like it takes like a good kind of month to collect like the right customer feedback. And so this is one way of accelerating which way am I going to go what questions am I going to ask so I believe in um a what I call a decision first kind of framework. You don't talk to someone just for talking to someone you have a a decision in mind and then you say in order to make this decision I need the following the information and then you say where does that information come from is it through talking to someone I mean interviews are very powerful. You can answer a lot of those with that but in some cases you don't want the interview you don't want that it's not the right technique to answer it. What you want to see is what did they actually do versus what they say because there's some things that people maybe won't tell you because you're on a call with them like how uncomfortable that that image kind of make you feel or you know it's okay because here's the founder of the company and I want to be kind of nice and polite. So some things you actually want to see what they did. So you don't just talk to customers for the sake of it you you talk for the like sake of I need to decide on something.
SPEAKER_02Okay so let me ask you with the speed like you mentioned doing something like in June of last year. Yeah right if it was anything else if it wasn't technology or AI in particular that wouldn't be that long oh last summer right yeah but in this world it feels like dog years it was like last summer was completely different than where it is now my time scale is I have never experienced this. And it's just moving faster too.
SPEAKER_00It's just if you I mean it's been like a year some things but if I look at code I built like 16 months ago I'm like who did this? What's going on here? Yeah and uh yeah the the time scale is hugely compressed.
SPEAKER_02So if you're if you're interested in developing this this environment of intelligence about customer feedback and sentiment and things like that how long do you have before that's stale or expires it's no longer relevant. So what do you mean like how long until that specific customer feedback is in yeah because it's got to have a small shelf life because the customer that said hey I tried this a month ago it was okay you know that it's not the same project a product 30 days later because it's gotten better.
SPEAKER_00I mean it depends it depends right on the kind of feedback but so you know when someone is telling you something right there's always a a sub kind of message if they're saying um I don't know you know there's that quote right the uh from what's name what's Henry Ford right if everyone yeah yeah if I asked people what they wanted they would ask for like a faster horse. Yeah he's talking about the car. But the quote's incorrect I think of course they want like a faster horse. They're not going to tell you how to make a faster horse. The car is like a faster horse. So the feedback is good. You should ask people regardless of the technology solution. Exactly the job they're trying to do the task they're trying to get done yeah is the same. They might complain about you like you you have to look underneath that always because you can't just say oh well they don't know what they're talking about because the tech isn't you know whatever it's they do know what they're talking about. They're talking about their needs and their needs are always the same. Your needs are always the same they haven't changed. Yeah regardless of the tools that you regardless like uh they if they complain to the photographer look uh my mole on this picture what they're actually expressing is I don't feel comfortable with the way I should look and I need an image that represents me in a certain way for this context which is at a conference and I don't feel like there's an actual match between that image though and the context doesn't matter what the tech is it doesn't matter whether it's an AI tool or anything else the need stays the same. So I would say it depends on that. Some of the feedback that we have on our older models I don't incorporate that anymore. Yeah uh so if they say oh my teeth are kind of messed up it's like yeah that was an old model so I don't I I don't always import all the customer data.
SPEAKER_02But you would be looking at almost like a meta complaint or a meta need which is I want a crisper cleaner um more per perfect image of myself whether or not it's reality that's what they want and then you look at the tech to support that yes and no more like where is that image going to end up okay that's what I'm kind of thinking about because it's not really just for you and it can be some images we've done now like entertaining images.
SPEAKER_00Yeah we're just about to do like a face swap style nice so you can yeah uh put yourself in any in any kind of meme or whatever. Yeah uh but but in in in general it's like the customer doesn't tell you where they need the image and why there's a problem. So that's where you have to dig into that. The AI can't answer that sometimes because the AI is also kind of just using the text it's got but it can give you a the aggregate view of that can give you a hint of where what you should ask the customer about. Ultimately with our images or the photos they're gonna end up somewhere. Yeah that's what it comes down to right and so if they're complaining about something it's that it's not right for where I need the image. And that's what you have to try to to basically find out then is where what are you trying to do with this?
SPEAKER_02So I I got a final question here and it's it's interesting because it kind of ties into to the entrepreneurship. Goes back to the you didn't start a company in 2014 that was intended to assist in you know resume selection. Yeah you identified in your own use like oh wow it would be great if we could build something or if we had something that would do this for us and then you saw that it was a business. Now with this with images do you I mean obviously you have a uh an interest in it now because it's it's the business that you're building but was this an accidental discovery or was there some interest in solving this particular problem?
SPEAKER_00Are you sort of asking what like the core kind of motivation is that got that for for you of all the options that you've had you've had successes across different industries and some failures and some failures and but AI allows for you to really uh participate in places where you didn't have prior subject knowledge you've now got expertise through the LLMs or whatever how did how did this one um how did you land on this of all the options that you had so you know it's like a existential question that you're asking like when you're asking me my mind is exploding with all the like the sub levels of actual motivation like I the sort of the the the right at the core of that is how I see myself and I've always just seen as an entrepreneur a person who likes to build things I like to build stuff. I've built many things that nobody cares about and so I've actually learned to try and find things where hey somebody actually needs those that can it can actually help somebody but I think coming back to the beginning of your question which was kind of like the accidental part that is a sort of you're asking the the sub question there is like how do people come up with the the ideas? How do people find stuff to kind of work on? Yeah and and I think there's you know it's a hard question because there's okay look the one way and this hasn't really been for me up to now really it's been like there's this problem in the world and I by the way I have some ideas on this now that I'm thinking about now like okay there's this big problem it's gonna be a big problem in like five years and even bigger problem you know like in 10 years and then you just like I'm gonna build whatever I need to build to get to that end goal. And that's the sort of SpaceX pitch right okay that's the Elon Musk you know in order for us to survive we must make life multiplet multiplanetary and everyone can quote that back and and the beauty of that is it's a big goal. Yeah it attracts people it's almost an easier company. SpaceX is an easier company to build to some extent because it's such an ambitious goal. Whereas if you say oh I just want to you know create pictures of people right so so with me the accidental part uh it's just been the what's the curiosity I've always tapped into that and the one thing I have to turn off and if you want to build anything of actual value I think you have to turn this off is the kind of judgment that comes from you and the judgment that comes from the outside. So you have to be prepared to make bad work right a bad first draft a bad first app and you actually looking for things that show you what the potential is not external kind of validation. So this so that's where my ideas have always come from it's like oh wow I didn't know I could do this and AI was an interest because suddenly you could do something with AI I mean you couldn't do a picture of yourself in the past yeah and suddenly you can do this it starts to bring up questions well what does that like unlock and for example that's where the kind of fashion comes from right because if I can put myself in a picture and I can say what's in that picture why can't I try on something? Yeah okay you see and then maybe by the same token there's something else there like oh maybe uh can I see myself before and after if I go to gym for one year there was a startup that actually did that. Interesting. Yeah I I mean I don't think it's a it's a repeatable product that much but it's that's what it unlocks. So the trend with AI is sort of this compression of the task. So the tasks have always stayed the same right I need to go get a picture because there's I'm going to a conference it's always been the same thing like or I need to get a picture because I want my what's family to have it. So the way you would do that in the past you would get a a painter to paint you. Yeah right so it's like same same problem same thing. But not many people could afford it because it was you know a painter yeah expert like a so with ai we're seeing this compression accelerate in the field that we never thought it would which is like in the knowledge worker field. Yeah right we all went from well we don't have to work with our hands we're not gonna dig ditches we're not gonna work on the farm picking like the fruit the machines are going to come in eight years. Yeah and so this compression if I see that anyway if I suddenly see I can do something that otherwise would have taken me hours or days there's a huge opportunity there. And so like I uh I made a podcast which I think you made an episode about now you spoke about the the AI generated podcast oh the two hosts yeah I think is that from because I posted the one no actually okay yeah it just it kind of hit the scene a couple of weeks ago and oh okay one of the fellows in our community was like hey check this out but you did a post about the same thing yeah I made two of them so okay so for the people listening like what are they talking about there's a new Google product notebook LM notebook LM I I bought an open source version of that now anyways but um it's just combining different things yeah I'll I'll give you access to it but yeah um and so this Google product you upload a text file or a PDF you've actually seen it right yeah and then it generates a podcast with two AI kind of hosts and they're talking to each other and like us like they'll interrupt each other and they'll go yeah and they'll confirm and it's like that. And so the first thing that came to my mind is like there's probably a huge company here to be built in or at least a content company that is actually focused on taking the written form that is not currently available in the the audio form. So a very dense book about some what's a what's a some some kind of philosophical topic that is going to take you is like a lot of kind of friction to actually read that and creating an engaging conversational medium to translate that a particular message. And so I I still think there's quite a big company there to do that. I don't know if I'm interested to do that but that's so coming back to your like original question of like how do you get to this what what sort of drives you to actually work on this is I'm fundamentally what's curious about things and when I see there's a compression like suddenly I can do this and I couldn't do this that's what we call the like inflection point right there's some like inflection here we're going through a huge one now. So the huge like inflection point is AI. Yeah and that's going to basically cause a whole bunch of things that you couldn't do in the past you can do or it would have taken you many expos to do you cannot do it interesting. Like code right you're gonna be building your own apps in a couple of years. Yeah well I have to listen to this where can I find the this uh episode you were talking about uh I recorded two I called it based cast a i I just posted them like on YouTube okay uh one of them is uh a few choice essays I had from it's Paul Graham don't know if you know Paul Graham he's the you know like the Y combinator guy and um the second one was from a book called Courage to be disliked which is a great philosophy book based on Alfred Adler he was a psychologist and I extracted a chapter of the book and I fed it into then I made two episodes like deep diving into that topic.
SPEAKER_02I'm gonna have to check those out that's awesome. Yeah everybody thank you so much listen big fan of Sector as I mentioned if you look at our uh any of the marketing that we use and here's what I do now is in anytime I'm doing a presentation and I'm introducing myself and I do presentations all the time, I will, you know, you have the the hero shot and everything. And before I change slides there's an animation that pops up and that says that this image is AI generated. And the reason that I do that is because I want people to to kind of go wow yeah like question reality like oh my he did that with AI? How do you do that? Because with this podcast with the questions that I ask and all those types of things I'm wanting people to get curious about oh I didn't know it did that. Oh I didn't think I could do that with it right and it's constantly changing. So we're doing what we can to let people know that that this is possible and that uh sec does the jam for sure. Chris thanks for having me for sure any closing remarks for um everybody obviously please go without um any endorsement from Marco please go check this stuff out i uh you will get used for it and if you are out there and you particularly a knowledge worker or somebody who is presenting as a professional if you've got you know a crappy image that's the there used to be an old head head and shoulders commercial that said you never get a second chance to make a first impression and that is that is the first impression that people get it's not the the title that you've got it's not the banner image it's who is the This person, and if it's not uh a representation that you think, whether or not you've ever gone out and taken that picture, but that you think, hey, this is how I want the world to see me, um, it's not that hard. How long does it take to generate uh a gallery now?
SPEAKER_00Within an hour, you have everything back.
SPEAKER_02So I I would encourage you to look at these for sure. Um it is of all the modalities of AI that are out there, certainly people are generating great content, video is not there yet. This is one of those AI modalities that I would say is is there. I mean, like I routinely use these images and people don't know, and that's my intention anytime I'm using AI, if it's coming across as synthetic, it kind of defeats the point. I don't want people to uh see something and go, obviously that's AI, because to me, it loses any gravitas that it had, any any um, like I'm not gonna pay attention to that because it's generated by a computer. But if I can make it to where it's indistinguishable, or as my friend Jody Cook says, indiscernible from human, like human output or or a human, that's that's the intention. And I would suggest that if you haven't done this yet, Secta is nailing that when it comes to thank you. Um, for sure, man. I mean, like go to their website and I think you recently changed the homepage, but you used to have like the the kind of the grainy black and white headshot where you could do the slider and it would go from that into Marco in a suit. And is that still the home page?
SPEAKER_00Uh yeah, we've uh I'm I'm a bit tired of pictures of myself on the site. Uh-huh. So I've been pushing the team to create like but it's a great example, yeah. Yeah, and so that one actually worked quite well. So I think that one is still there, but we tried different images just to see which one like communicates it a bit better. But I mean it just maybe something quickly that you said here is that there's you actually mentioned something that was very interesting, which is the the authenticity of it. You said, like, I need a picture of me, and the people that will push back to you now on everything AI is the what's the purists. They're like, okay, you know, it's not the same. It's not the same if an you know a photographer does it, and there's just that human touch and the sparkle, and there's and I understand where they're coming from, but the purists have been around in every single technological change. I mean, there were film camera people, right? Who were like, how dare you use a digital camera? Yeah, film is where it's at, like film is pure, and even the what's in the film, you know, the directors, like the movie that directors were with film cameras, digital cameras, AI. AI is just like another tool, it's like a replacement for the camera. So, yeah, I think if you get rid of that suspension and you keep playing with these things, that's where you'll get the most uh actual benefit.
SPEAKER_02Awesome. Um, any closing remarks or anything that people should be thinking about on this topic as we wrap up the episode? Keep playing with things. Yeah, 100%. Everything will come out of that. Awesome. Well, thank you everybody so much for uh joining us for this episode. I'm excited to continue the conversation off camera, but um, thanks again. And uh Marco, thanks. Uh this was something I think that's been in the works since the first time I did my SECTA uh images images back in August of last year. So thanks again, bro. Hey Chris, thanks so much. Thanks, appreciate it. Bye-bye. Thanks for tuning in to using AI at work. Don't forget to subscribe for more conversations about how to use AI at work. And a special thank you to our sponsor, Chief AI Officer, for empowering businesses with AI education and training. Visit their website for a free AI readiness assessment and AI strategy guide to help you get started using AI at work. That's www.chiefaiofficer.com. Well, thanks to our producer, Evan Desponier, for making this episode possible. Follow us on Twitter at the handle usingAI at work and visit www.usingaiatwork.com for free resources to help you harness AI in your role.