This Broke the Internet: The ChatGPT Update That Changed Everything

Taylor Holiday: [00:00:00] Welcome back to another episode of The Ecommerce Playbook Podcast. With me today is not DTCJacob. He is transformed. He is evolving every day. He is now Jacob_Posel. And he is you think that's where you're gonna land in final or is that gonna evolve again, do you think?

Jacob Posel: I mean, if I can get like @Jacob or something, that would be sick. 

Taylor Holiday: I mean, that's what I'm, I just …

Jacob Posel: yeah, I think, I think I do have Jacob Posel, but I, I think it's like an old account that I don't have access to anymore, which sucks. So like, I'll settle for the underscore.

Taylor Holiday: you actually do own your own name, but for now it's an underscore, so you just like can't log into it or something.

Jacob Posel: Yeah. Yeah. I have an old previous life of the best friend of an influencer. So I used to have a Twitter account, but that's like a story for another day.

Taylor Holiday: Oh man. Well, maybe you can get that. Yeah. You'll just keep sort of shorting it, evolving it until it's just Jacob eventually. So, does, who, is there an at Jacob? What is that? Jacob? Is that

like who, who who currently possesses this name? I.[00:01:00] 

Jacob Posel: I'm not big enough yet to own at Jacob. I don't deserve it yet.

Taylor Holiday: J Jacob, Jacob Mullins, venture partner at Village Global.

There you go.

Like maybe that's can be part of your investment thesis. Is that like

you get the 

Jacob Posel: out at Jacob. Yeah.

Taylor Holiday: anyways. Sidetrack. But we are here because Jacob broke the internet this week. As a result of the newest chat GBT model update, which sidebar before the podcast started, Jacob had referred to as the, the biggest model update that he had experienced since going back to maybe chat GBT three and a half to four.

But we're gonna talk today about specifically the impact that it made on ad creative for e-commerce and what he discovered in that process. So, Jacob, take, take me back. You were watching the live update.

And you see them announce this

and then you just go into motion. What

did, so what was the aha moment for you, and did you know that it was gonna get as much attention as it ended up getting?

Jacob Posel: Well, I mean, to really start the store, we have to start like six months [00:02:00] back when I.

Taylor Holiday: Right.

Jacob Posel: When I was working, working for CTC and like CTC is always early on things. I feel like that's something that I'm starting to learn. We were working on something called the Performance ad Library, and so I was always trying to solve.

AI ads, and I couldn't get it. Like it, I was defeated. I, it like, I just couldn't, I just couldn't figure it out. I couldn't get AI ads and I didn't think the technology was there, but it was like always this thing in the back of my mind, like it was like my pet project that like the thing that defeated me was the concept of AI ads.

I, I couldn't build it. And like, honestly, no one had really solved it yet. People had tried, people had like gone viral trying to, but no one had solved it. So I always watch the open AI product launch live streams because like the world can change in after one live stream. It's like that's how, that's where we are right now.

So I was watching it and I saw like this. Massive improvement in image capability in like product [00:03:00] generation in text generation and placement. And so the very first thing I did was I try to make AI ads. And I didn't add for ridge because, I dunno, it's just always the first thing that pops into my mind.

But it just worked and I was. Blown away because this thing that had been so unattainable is now created. And I posted on Twitter, I said First ads created with chat GPT. I. And I didn't know it would blow up like that because it, like, over, over the course of the week, I just like spammed Twitter. Like I really went ham.

And I, I made a lot of things. We uncovered a lot of cool things and it just went absolutely crazy. But yeah, I think it, it like started a wave that kind of broke the internet and it was just so incredibly impressive.

Taylor Holiday: So the funny thing is, so you're right, the, the, the frustration we experienced, I think together, and I run into this a lot for myself, and I, I get feedback that I often believe the future is too close to the present. Like I, I, [00:04:00] I, That's a hard thing for me to distinguish and I feel this a lot with AI right now, is that I don't really know how close my feeling and idea is to reality.

Like they feel really disassociative in some way. I think AI creative has been that where you and I had been working on this project for like six months where we were both like utterly convinced that this is going to ha that, that this is like an inevitability. It is just for sure going to happen, but it would just like, there were all these hurdles where. It just hadn't crossed the chasm of people's mind to believe it yet. Like they, they just, they just didn't believe it yet. And, and it's pho because so much of it revolved around text, which just felt like the, in my head, not being an engineer felt like there's so much structure around font libraries.

Like, how is this the problem in this like thing, you know? But yet it was such a massive hurdle for people to conceptualize because. That sort of font text treatment is such a big part of advertising. Like it's such a, a thing that makes people feel right about it. And then there were little subtleties about products and shapes and things that would just so [00:05:00] we could never get to a place where you were able to do it where. You got some brands to like the assets. We did some stuff with a polo company and a few others where you could get there, but it was like laborious,

right? Like it, like you had to really work at it. And we were talking about like 360 photographing images and trying to get all the best training data and blah, blah, all that.

Like to, and this is like the part that I, I'd be curious like, to think about is that where do you like trying to figure out how to enter this stream when it's moving this fast is so freaking hard to figure out.

But anyway, so you, you see this, it happens and. Was it like, did you, was it obvious to you that this was going to be the moment that everybody believed?

Because I, I think that was, it's been hard to figure out is like, why don't people believe this is happening? Like,

it seems so obvious, but when you first went into chat GBT and not watching the live scene, did you know that, okay, everybody's gonna believe this now?

Jacob Posel: I didn't know because creatives like, particularly, particularly on Twitter, like the worst people in the world, no offense, like I hate interacting with them. Sorry guys, but like, I didn't, so they're always picky and [00:06:00] they are afraid of like losing their jobs or whatever, or like they always will find a reason to dislike it.

But I was hopeful because I felt like it was no longer deniable. You could no longer deny that this was good. Like it, it just was objectively good, you know? 

Taylor Holiday: They still will though, like they still will deny that it's good.

Like Barts things like I, it's a funny like, but anyways, but yeah, they'll never stop. Some, some subset of are unviable, but, but you're right.

Jacob Posel: yeah, yeah. I mean, there's a fine line between. Between Madness and Genius, you know, but like, I think the, the majority of people realized that this was really big opportunity and it was like so much fun. The vibe, like the vibe concepts had finally been brought to creative and people really enjoyed that.

Like the ability to create magic at your fingertips was something that I. Was missing for marketers, I think, and for creative people. And now it was there. And so like it was accessible, which was, which was really nice. Like the previous generations [00:07:00] of image generation were not accessible and it suddenly was, and I think that was part of the magic too.

Taylor Holiday: Yeah, for sure. So what, what kind of crazy things did you experience in the process? Going viral on the internet, like how many pe, anybody? Cool. Reach out. Anything fun come out of it that's been surprising to you?

Jacob Posel: I had like some really surprising stuff. Like I had some people from the phase clan reach out to me, like the, the online group. I had some famous, like a really big YouTuber from England that I used to be a big fan of, reach out to me, which was like kind of weird, like people unrelated that you wouldn't think of with millions of followers.

But for the most part it was a lot of people asking me to create images for them. And I was like, just go to cha PT. Just go Please just do it. 

Taylor Holiday: it is like you're missing the point here. You're missing the point of what I'm trying to show you.

Jacob Posel: Yeah. Yeah. And I was posting tutorials on like exactly how to do it, and I would still get people saying like, you're lying, you're lying to me right now.

This wasn't Chad GPT, and I'd like 

Taylor Holiday: Yeah. 'cause it, it was funny, like you would see that in your comments where people would be like, I'm trying it, it's not doing the thing you're [00:08:00] saying you're, you're, it's, it is, and it's like, no, like literally I'm showing you, this is the screenshot of what I posted. You could

try it for yourself. So now that you've seen this here, okay. I have two questions. One is. It feels like the next big step here is the API release coming up. What do you know about that in terms of how that's going to impact this process and how are you thinking about the interaction as it relates to when this becomes a. More of a resource that you can leverage, because right now the process is still sort of very one-off. You get an image, you

make one ad, you get an image, you make one ad, you give

a prompt, you make one ad.

Where do you think it goes as that portion gets unlocked?

Jacob Posel: Okay. Bit of a bit of a first for, for the CTC viewer, the Econ Playbook viewers, because this just happened. But I saw some people who have been able to reverse engineer the API, so I'm talking to someone there to see if we can get some early access. But that is gonna change everything. People are [00:09:00] still, people are still saying like, oh, but you could just do this with Canva.

You could just like create, you could just do it with Canva, but it's like. You're gonna get overwhelmed by volume to a place where you can no longer, it's impossible for you to compete. So we've found that the best patterns and the best way to do this is like, it's usually to provide an inspiration image and then tell it to make adjustments based on like your product or something.

And it does a really good job of taking that inspiration, combining with your product image and creating an ad or making adjustments to your existing ad, something like that. And so what I'm gonna do, or what people are gonna do, is they're gonna take an inspiration image, and then they're going to take every single product image you have or however you wanna do it, and they're just gonna run that in parallel for thousands of ads, and you're gonna like.

I don't know exactly what the interface looks like, 'cause it's gonna be early, it might be like a Tinder swipe thing. You might have like an AI evaluator, which takes out the the bad performing ones. But you're gonna be able to do, you're just gonna be overwhelmed by the [00:10:00] volume and you're gonna have so much choice that you're just gonna be able to pick the winners.

And so it's no longer gonna be like, create one ad and then make a little adjustment and then hook for another one and 30 minutes later you have five. In 30 minutes, you're gonna have like 5,000. 

Taylor Holiday: So, yeah. Again, this, there's sort of like the infinite possibility here. Do you think there's gonna be, so it seems like a couple things are happening. One, it, it appears that there's some constraint that's happening with chat DPT as it relates to the demand versus expectation for this process. So what do you think?

Do you think there's gonna be limiters that are going to prohibit that kind of behavior? What do you think is actually gonna be the limitation to that coming to life quickly?

Jacob Posel: I mean, it may be expensive to generate an image to the point where people are still gonna have to pay like five bucks for an ad or something rather than what they were usually paying for. That may be something that people need to consider. I. It may be a little bit slow, but when we're talking slow, we're talking like a minute, which is like not that [00:11:00] bad.

The,

I think that the only thing slowing down the API, right, API being released right now is like the number of GPUs that OpenAI has. I think they're already struggling with just the consumer version of this. And when people are able to go crazy, like what I described, there's gonna be a ton of load for their like computation abilities.

And so that's gonna be something they need to figure out. And so it may be rolled out slowly. OpenAI has different API tiers. Lucky CTC is on a pretty high one now because we use AI stuff. But yeah, it's gonna be, there's gonna be ways that they are unable to handle this, but like it's absolutely inevitable.

Like this is the, this is the easy stuff for them to solve.

Taylor Holiday: Yeah. Yeah. And that'll just keep, keep becoming faster and cheaper, repeatedly, endlessly over time.

Jacob Posel: A hundred 

Taylor Holiday: So, 

Jacob Posel: no time.

Taylor Holiday: so as I think about this, like, it feels like even the intermediary of, like you said, the picking, like I. It feels like even that is going gonna go away. Like for some [00:12:00] subset of brands, they're actually going to eliminate their own need to intervene between the ad creation and the delivery itself, where there's just a campaign that

is endlessly updated at all times

by ai.

Like that to me is the end conclusion here, is that like the whole delivery system from ad creation ideation into delivery, into optimization is just done 

right? Like it's just over. 

Jacob Posel: Yeah. I mean it like this doesn't just change creative, like this has to change. Media buying too. Like the, it, like the thought experiment is what happens to media buying when there's zero constraints to the number of ads that you can 

Taylor Holiday: That's right. That's right. 

Jacob Posel: I think that's where it's heading. And like, I don't really know what that means.

People are gonna have to figure that out because it's been three days. But like, that is the, the way that people need to be thinking about this. It changes everything about, 

Taylor Holiday: That's right. Because in so many ways that is, right now, the constraint that drives the entire decision making and logistical process of media buying is when am I getting my next ad? Then I will upload it, then I [00:13:00] will take action.

And if that just goes away and it's like you have an infinite pool of always available stuff, what would you do?

Like that's a, that's a, that's a question set that I don't think has been explored very well.

I, I, so do you have a point of view on that? What do you think it does? I.

Jacob Posel: Oof. I mean. Everything has to change, like from the way that employees work to like the way that, like, I don't even know if how the auction systems, like, I don't know if they change as well based on like the inputs that they receive. Like everything now works on a monthly cadence basically, or like a fortnightly cadence.

Like that's no longer gonna happen.

It's like. It's like what? What then becomes the important piece and like what becomes the structure of your account where you no longer have to do that? I mean, the only constraint there is can you get them uploaded, but we know that's gonna be solved as well. So, yeah, I mean, you're a better person to answer that than I am.

Taylor Holiday: Well, I, I think it, it, it's so interesting. I, I, you're doing an interesting thing on Twitter right now that I do think sort of relates to this, which, and I'd be curious about this part, which is, I, [00:14:00] you tell me if I'm wrong, but you seem to be attempting to illustrate. Or fight against this idea that like, sort of the cri the, the like lowbrow criticism of AI creative is, is like it's all duplicitous and rep.

It all sort of moves into sameness and it's not really creative and interesting. And I get the sense that what you're trying to show is like, no, no, no. This is like the unlock to infinite creativity. Like it

widens creativity. It makes it. Yeah, you're gonna be able to to do things that have never been done before and never seen

before more than ever. And so how do you think about, but that does fly a little in the face of what you described before, which is like, here's an existing template, here's my product replicate. That's still more copying than it is net new ideation and novelty. So how do you think about the process to generate. That new idea.

'cause it still seems to me, you can tell me if you're, you're generating the idea for, what was the one I was just looking at? The, the woman drinking monster in an old school [00:15:00] Panasonic camera view. Like that you're, that

you're making, or the woman on a mini bar holding the spilt over Michelob Ultra.

Like that's still Jacob's brain. Right. That's generating that ideation.

Jacob Posel: Yeah, it's like, first of all, those two, I have to give credit to the actual original. I, I was kind of trying to you can see in the comments I gave credit to soa, but those are from the 

Taylor Holiday: Oh, okay. Yeah, yeah, 

Jacob Posel: But there are some other really creative ones that I tried to put together. Like, for example, there's this legal startup called Harvey and I showed like Harvey Specter with his brain half open with a robot, with like the Harvey Chip in it.

Taylor Holiday: Yep. 

Jacob Posel: I, I just truly don't, I, I don't think it's correct when people say that this, like, I. Everything is gonna converge into sameness. That's not true at all. Like you can create ultimate creativity here. There's like, there's just multiple different elements to it that I've been discovering. You know, like if you want consistency and reliability and a part of this is just me trying to kind of, I.

Work with the people who are criticizing it. It is to take existing [00:16:00] templates and inspiration and create variance of those because I know that's kind of what people will approve of already. But then there is this other way that you can create true, true creativity like the Harvey one or like the MCC Club Ultra one that you're seeing as well.

And so something that I've seen is like. G, PT four. Oh, it's, it's a multimodal model. Okay. So there's like, there isn't this concept of an image model and there isn't an a text model. It's a model that understands both, and it's able to express its creativity on the images. It's not just. It doesn't need a very, very strict prompt like you'd need before.

And so I think AI is gonna be both the creator, like there's gonna be human creativity in there as well, but it's gonna be able to express itself as the creative and as the generator as well. And that's really exciting in its ability to do that.

Taylor Holiday: heart. Right. And, and I think it's no different. Like the way I think about it is even in the, in the process of designing the system for the scaled creative output, you could, you can build in ideas like creative diversity, like [00:17:00] novelty, like, wildly diverse visual aesthetics. Like you can,

you can essentially program in that output, right?

In a

way that does. Yield that kind of result. And I think I, I think that you and I have always sort of talked about this is that like, what is the underlying desire that you sort of build into the prompting of the system That does still matter. But my question though is like, is there, is there a intermediation of that though that allows for the feedback of performance data? To entirely drive the creative process, absent any human connection, and you're just simply stating Produce X outcome. Ready? Go. Like, is that the end output here?

Jacob Posel: I think that's where it can go. And I, I think that like the foundation structure for that is being built like Compass and like other like creative strategy tools, whatever. I think that like, I think that is what is going to have to happen. I. I think that in every [00:18:00] step in that journey, the you, you can, you can give your goal and you, you have to be able to define that goal, but once you give it, you have to allow AI to kind of drive towards that.

And now that it has the biggest. Like the, the biggest bottleneck was the creative. And now that if, now that that's solved, I think it can do everything else really well. Like we've seen that it can analyze data incredibly well. We've seen that it can call tools and we've seen that it can do all, all, everything else that you need that a human does.

It's now, now that you've got like, it's like the fuel to, to the to the engine. Now that you've got that solved, I think that is where this has to go.

Taylor Holiday: What do you think happens when there's AI systems competing against each other in the meta ads auction? This way, like

Jacob Posel: I mean, like. When technology changes this much, is this even gonna be a SA SaaS tool? Like, like, like let's just say CTC solve this. Like [00:19:00] would they sell it or would they just become the best marketing engine at the cheapest cost in the world? Like, I don't know. I don't, I don't know if that's sold. Like for example, a hedge fund, right?

Like if a hedge fund figures out an algorithm to solve the market, they're not gonna sell it at a cost. They're just gonna solve the market. So I don't actually know what that's gonna look like.

Taylor Holiday: Yeah, I think, yeah, it's a, it's an interesting question. It's sort of like one of the things I've always taken a little bit of a. The, the Shopify market position has always frustrated me a, a ton in that like, I feel somewhat gaslit by the idea that we're building we're arming the rebels in, in this like way that's like basically to the detriment of everyone because it creates homo homogeneity around the execution using the tool.

Like

if we all get the tool, then my, I've gained nothing as a, as a user.

And so there's this question for me of like, when I see something like this that I'm always like. Anytime we as an industry go, oh, this is so exciting. I go, oh, no, hold on. We, we just like introduced. Like sameness [00:20:00] in the accessibility of it, like the access.

And I've heard this argument as it relates to ai which is that it's like it's not a value creator, it's a utility. Like, it, it, it moves to the level of electricity in the sense that like there's, there's not this like value to be captured off of that, the thing itself as much as what is done with the thing.

And so, so I'm just curious, like, and, and this is sort of now getting to Jacob's personal life and your sort of quest to go and build. The next great thing, like where is the value capture that, like what? What do you think about this?

Jacob Posel: I am like. I actually don't know. I actually don't think that I have the answer to where the value is here. And so like, I'm it's not necessarily where I feel like is the highest conviction area to build, but I can say that I think that there's this assumption that like everyone gets AI and so everyone is now able to do is, [00:21:00] is able to do this, but like from what I've seen, that's not true.

Like. Chat, GPT as a, we can just use text because text is like that. That's probably the that's probably the slope that image is gonna follow where chat GPT text Generation has been around for. I. Has been around for like two years now and it's really, really good and it can, it can basically do anything that you need a text generator to do.

And like people still dunno how to prompt. They just don't like they, they still dunno prompt best practices. They still dunno how to get the best out of it. They still dunno how to use it to the maximum ability. And like, I don't, I don't think that image is gonna follow that any differently. I think the highest agency and the smartest people are going to figure out ways to use it to maximize their.

Like, maximize their goals or whatever and like use it really well. And I think that's 1% of people and I think 99% of other people are too lazy or too slow or aren't able to do it. And I think the same winners are gonna continue to win. Like that's, [00:22:00] from what I've seen with ai, I just think winners will continue to win.

The people who haven't, weren't able to solve it before, I'm not gonna be able to solve it now. And I, I, I, that's what I think will like ultimately happen.

Taylor Holiday: So you don't, you think it does still concentrate into some human skill? Like it that there is still some human skill required to maximize the value of the technology that is going to be like other things in that way.

Jacob Posel: I mean, there's multiple steps to adoption. Like I, I think that particularly in like the e-commerce world, I think that if I built. The best tool in the world for doing this. I don't think that everyone would adopt it and use it tomorrow. Like I, I just don't, I think people are too slow. They wouldn't take the meeting, they wouldn't take the call.

Taylor Holiday: So then how, well, but how do you, what about icon me, who claims to be the fastest growing company in the history of the universe or whatever they claim, you

Jacob Posel: Yeah, I mean, how many customers did they get? Like a thousand, 2000 or something? Like, I, I, maybe they're growing really [00:23:00] fast, but how many of those people signed up and how many of those people started using it every day? I bet you that was like a, a very high drop off, but icon. I mean, I, I, I don't think, I don't know if that's necessarily means that it's saturated the entire market, like assuming it can do what it does.

I don't think that it's already saturated through the D two C market.

Taylor Holiday: Yeah, I mean, I don't know about saturated, I don't know what the threshold would be, but it, it is interesting to me that. When you get this like marginal step function in change in price at a, at a moment when the market needs it so badly, like if we, if we go and we assert that like the cost of ad, like if, if there was a way to sort of calculate the cost per ad production across e-commerce right now to say, to make a high quality stack static image ad on average costs x, let's say it's $150, 200, whatever, it's video 400, whatever, and then all of a sudden that becomes available at, you know. $200 a month for infinite choices. Right? I think that you can drive in the short term a lot of adoption [00:24:00] for that, like really, really fast. The question is, what happens to that price over time? Does it, does it just keep going down such that, or do the base level things that it's built on top of like what Che CBD does?

Do they, does it have staying power? So there's sort of that, like what is the sort of, the long-term tailed value of the tool. At any moment, given that step function changes to what Chad GBT seems to be doing, which is like they seem to be just saying, well, we'll just do all of it,

Jacob Posel: Yeah,

Taylor Holiday: you know.

Jacob Posel: this is my, this is like my actual kind of projection or forecast of what I think will happen using GPT-4, using kind of like what happened with text. I think that

GPT-4 oh all the way to like oh one and oh three. GPT-4 was a, was like a massive magnitude of magnitude change and then it started linearly increasing.

I don't think we're gonna see a, this level of change in image generation again for the next couple years. So. I [00:25:00] think it's gonna stay where it is now, and it's maybe it will get slowly, incrementally more better, and then people will start building workflows around that and offer that as a service or as a product for people to start using it and adopting it.

But I. It's not, a lot of people will want it to produce the perfect ad every single time with the first prompt. Right? And we've spoken about how like you're gonna be able to use volumes to get over that and like, that's what people are gonna try and solve. But I don't think that it's gonna be the perfect kind of.

Solve all solution for every single person, every single time. And I think that makes a lot of people churn off of it. And the peop again, I think the winners will stick to it and like figure out how to use it and how to adopt it and maximize it for them. But the other people, because it's not gonna work perfectly every single time, they're gonna stop using it.

I think given a world where everyone had perfect motivation and ambition and agency, I think we would see the world world that we described where you can go, everyone just [00:26:00] puts ev perfect ads into the meta ad auction and see what happens. But I just don't think that that's the reality because it's not gonna be the golden bullet every single time.

And so people will turn off and the best people will use it.

Taylor Holiday: Yeah, I, I think there's also one of the things that tends to happen is. There tends to also be like a, you can actually win this battle as the, as the technology layer becomes more homogenous amongst competition, you can win with ui. Like in the sense that there's gonna be some way in which you described it, like, I think the example that you gave is like dating apps won on you know, the sort of this hinge style swipe left, swipe right.

That became a UI change that wasn't a technology change that altered human consumption of the information. And so I, that's something I'm thinking a lot about is like. What is the actual like delightful, marvelous way in which you consume all of this creative and get it into meta and get feedback to it?

Like there's something that I think is gonna be a big change there that is going to be just the best way to do the thing that everybody's gonna be able to do, but it's just gonna be the most [00:27:00] delightful, simplest, enjoyable way to actually interact with it.

That I also think is an interesting yet to be solved problem is that I just don't know what the interfaces are for all these things yet.

Jacob Posel: Yeah.

Taylor Holiday: So I don't know. I, that, that, that, and this is my big problem too, like right now, I mean, you, you, I don't know how far to go in trying to build product on top of this stuff right now. Like, I just don't know where the value creation lies. Enough to feel confident in it. And even I think I get the sense that you're sort of figuring the same thing.

It's like, where do you enter, where do you throw your, where does the, the sign show up on your door that is a business name and what does that business do? Like, I'm still waiting and it, it hasn't, doesn't exist yet.

Jacob Posel: Yeah. Yeah. I mean, I think that there is gonna be a way to make a lot of money very quickly in doing this thing like a, maybe like a short spike. But, but I am unsure in where the, I. In, in a way like the long-term value is created here. And like there may be second order ways to, to take advantage of it that [00:28:00] I'm not seeing right now.

But like I would say that the best way to figure out how to create value is like figure out where it works really well and where it doesn't work well and just kind of focus in the short term and where it does work well and don't try and solve where it doesn't because that's really difficult. And then kind of build workflows around that.

But like. Everyone's gonna get the API at the same time, you know? So like, I guarantee you motion is gonna have one and icon's gonna have one and everyone's gonna have one. So like, that isn't where, that isn't where the value comes from. You know, it's not, it's not doing that. API thing that we discussed.

Taylor Holiday: The, the part that I actually think about a lot is I think the value is in being the kind of brand that's willing to, relinquish this like brand standard premise that you had and your design approval process in a way that brings freedom to the leverage. Like, part of me wishes I was starting a retailer that was almost entirely driven from product selection to distribution to ad creative from an AI level.

Like to me, there's like this fundamental root change that actually happens at the, like [00:29:00] consumer buying level that alters how all of the service and brands off the back of it change that, like that, that's where I feel like something's gonna happen first, whereas we're trying to like. Use this really, really high powered machine in these like really old structures and it's like we're gonna make our design supply chain faster.

Feels like the wrong solution to this problem right now. And I just feel like that that's the piece I'm waiting to see what happens that I just, it's not clear to me yet

Jacob Posel: Yeah, I mean, the market is gonna decide. I, I also don't know if like the. The consumer actual, like a lot of people are saying like, oh, this, like this wouldn't sell. Or it's like this, like this one doesn't work. It's like,

Taylor Holiday: That's dumb. Yeah.

Jacob Posel: It's like at the end of the day, the consumer knows. The consumer's gonna decide how it works.

I think like the best thing to do is just to.

Be close to it so that you can take advantage of these things when they come and be like agile and be quick about changing your workflows and your processes around doing it. And don't necessarily be open to, to revolutionizing and [00:30:00] changing how you approach things, and don't always try and fit it into your antiquated system, which wasn't designed or even thought of when this was introduced.

Taylor Holiday: Yeah. Yeah, exactly. So you probably have a lot of people asking you, right, brands, whoever are saying, okay. Other than like, I play with it a bunch. 'cause I think there is a choice where there is some legacy system. There's, from what we know, there's a designer, there's briefs, there's approvals, there's uploads.

Like there's this whole chain. Of work that's happening right now, and I think it is really hard for people to not think about editing the current workflow as the, the thing that they go to do with this tool. Like how would you actually, in help people to think first principles from don't think about your current system, imagine something different.

Like how, how do people do that? What is the process that you would actually enable for that? Would you create a separate, distinct workflow and not actually allow it to touch your current one? Like what would you do?[00:31:00] 

Jacob Posel: If people can afford to do that, then yeah, I would absolutely do that. Because like, I'll be honest, I don't think most people have optimized systems already. And, and so I think to fit this into your system is like not the answer. If you can think first principles, what am I trying to achieve? I'm trying to sell a physical product and how am I doing that?

I'm trying to convince people to buy something based on the ad that they're seeing, and I now have a completely different way of doing that ad. Like, how would you. How would you achieve that? Okay. So it's like, yeah, like you have, you had a designer before and you had to create a strategist because you need to think of an idea and you need to come up with a and come up with a, an asset.

But like, are those two things now necessary? When when charge GPT both understands the the process of coming up with an idea and understands how to create an asset for that, it's like, well, that, that's now merged. You no longer have to have that pipeline of one to the other, just like. Merge those things together.

That's like, figure out, figure out where things are no longer needed and no longer required. And try and optimize towards [00:32:00] that goal without try, without like fitting it into something that isn't like designed for it, I guess.

Taylor Holiday: Yeah, I, I, I think that that, the way I've been thinking about this is this guy I'm on the verge of, like the other day we were in the office talking and I'm like, I'm gonna ha, I'm gonna. I put out this like post on Instagram and I got a just basically saying like, Hey, do you want your career to transform forever?

Like, email me and I have an opportunity for you and I'm gonna put 10 people in this conference room, and I'm gonna say, you all have one week. Here's all the tools. Here's all the information. Here's all the people to follow. Whoever makes the best ads over the next 10 days, I'm gonna give you $150,000 job ready.

Go see you at the end of it,

because I don't even, because my part, my problem is I don't even know what system to try and start designing. 

And I, what I, what I feel like, to your point of like, I need the end output. I need a lot of amazing ads my clients love,

but I don't even, like, I don't feel equipped to go, okay, we're gonna go do this step and then this step and this step.

I wanna give people like a blank canvas. Tell them to go read the world, consume your content, go play with it, and just see what they did. See what they develop out of that space completely [00:33:00] unencumbered by whatever my current workflow is. It's almost like I need to learn a process more than I need to use the tools to change my thing.

And so I'm trying to figure out what is the best way to create that kind of environment inside of my company. And I'm tempted, like when I first started, saw this all happen. I went to a Lean and Nisha and all the people, and I said, go look at this and talk about how it would end the process. But what I experienced back was like, they started talking about all the hurdles to how the current process wouldn't work. With this, and I went, oh, shit. Like, of course, that's, that's their job. That's the expectation of the thing that we've built around them. But, but so they can't, they can't actually view it fresh enough yet, you know, like there's this limitation that I've created by putting them inside of this system that it's like, how do you create a space for a completely unencumbered view of the outcome that you want?

Jacob Posel: Yeah, I love that so much. Like, please tell me if you do. Like, that's like, sounds like the coolest thing in the entire world. I, I think like, that kind of has to be how you, [00:34:00] how you do this. You have to be willing to take the risk and to just be like, I created, I, I invented a business in a world where this didn't exist and this does exist 

Taylor Holiday: Yeah. exactly. Yes. 

Jacob Posel: How do I now, how do I now adjust to that world? Or how would this business exist given that this world does exist? And that sounds like the coolest thing in the world. I, yeah. I think that people will come up with very novel very novel solutions to it and figure out ways to create great ads.

And I think it will be, it can be completely different to the way that like concept to design, to add into media, buy into ad account works today.

Taylor Holiday: And, and like, I think it opens up this whole world, like I've felt this, where I have never been technically gifted at any of the technical tools associated with design. And I, so I've always felt designers always felt magical to me 'cause I could express an idea in words and they could bring it to life.

And that felt like they possessed a superpower that I wished I had because I just didn't know how to go in to draw the lines. Right? Like I just wasn't good at it. And. Now I feel like all of a sudden, if I was 20 years old, I [00:35:00] could redefine myself as a creative in a way that I never felt like I could before.

And so part of what I wanna do, and the reason I like sort of wanna put this broadcast opportunities out, is 'cause if, like, if you're a real estate agent or a, you know, whatever job you're doing, all of a sudden there's an idea that you could be a creative person.

Like, that's like now, right? Like yeah.

That's now in the realm of possible career options. And I think that's what, like, what I want is just someone who's sort of like, I'm dissatisfied with my life as is. I'm willing to be obsessive about learning this new thing and play with it is actually the human I need.

I, I actually don't want a designer with preconceived notions of Canva.

I don't want them to know what Canva

1. I don't want them to have ever heard of Adobe Illustrator. Right? Because they're gonna default into trying to think about. How that old workflow overlaps with this new one.

And it's sort of like, I don't, I just want you to come in and go, okay, here's these things.

You can talk to it

and it'll make pictures ready. Go. You know, like, and at the end of it, I will pay you a ton of money to become this person. Your whole life can change if you become en engaged in this 'cause that's what it feels like has the power to do for people. And [00:36:00] so, I don't know, there's something about that that I also like, want to unencumber us from anyone who has existed pre this, you know, in that design world.

Jacob Posel: Yeah. First of all, this is how you win and this is why you continue to win. So, I, I, this, that to me sounds like the coolest thing in the world. And like related to that, I was so bad at art in school, like in primary school, which is elementary school here. Like I literally failed my tea when I was like 10 years old.

My teacher pulled up my artwork and graded in front of the whole class because she didn't wanna fail me. So like the class was like helping me out to get me to like a 50%. Like I was that bad at it. And like, go scroll through my Twitter now and like, tell me that that is not creative. You know, like I have a completely different ability now that I'm so like not talented at because of this capability.

So yeah, it's like, it, it changes, it changes creativity for all people. And like, I'm, I'm like the perfect example of that. Like I'm the guy who went viral for it that literally can't draw stick figure.[00:37:00] 

Taylor Holiday: Yeah, that's right. I, I, I, and I think, I think inviting people into that reality is part of the beauty of what the tool can do. And so, hey, if you're listening and you wanna join my battle royale, please hit me up and let me know because I, I, I just, I, I, I am endlessly racking my brain about the right way to interact with this kind of seismic shift.

And I'm, I'm just really sure it's not like, edit my system where we work in Asana and create a brief and then ship it to some designer and they send it back, and then somebody approves it and then you like, it's just like, no. That's not, that's not, that can't be, that can't be the thing here. So,

Oh, what else, man?

What else do people need to know or follow? Or like, what, what are you, other than like, who are you following? What are you reading? What are you doing to stay up to speed on this as

it's happening? Because it feels like it is still changing so much.

Jacob Posel: Yeah. Yeah. I mean, the next thing that I think people should be aware that might come here is video. So 

Taylor Holiday: Yeah, that's gotta be, yeah. 

Jacob Posel: yeah. Be prepared for that. It video may be it may be have reached its peak with Sora. I hope not. And I think that OpenAI seeing the effect of [00:38:00] image will motivate them even more to improve video.

So like, that's something that people should be aware of. And like I just, I say this all the time, but I think people really need to just. Kind of do what, do what you're saying and just like get as close to it as you can. Like Twitter ultimately is the place where this exists right now and where things are moving the fastest.

So I would make sure that you're involved there. There's people who post really amazing stuff there, and it's like even the small accounts are uncovering like really, really cool, like new like best practices and stuff, which I never would've thought of. So it's like, it's just an amazing place to be.

And like, listen to the podcast. Like I, I think that you have to get obsessed with this stuff and like, I don't actually, people ask me this all the time and like, I don't actually follow anyone specific. Like, I don't subscribe to anyone specific, but my social media has become it's just AI stuff now because like, I, I just engage with that stuff and I really love it.

And I, so I think like, just being obsessed is the best way to get good at it and get close to it now.

Taylor Holiday: and what's [00:39:00] so cool, I think about X in particular is. There was a conversation, you and I were going back and forth where it's like if, as you encounter a challenge, like can we create a layered file out of this,

or like, it's like it solves itself like in the comment stream where

people grab the problem, they source the information response to what they're doing, and so it becomes this like, I. Hive mind of collectively improving

knowledge. That I think is like, why you do just have to be present. Like it's a weird thing to describe to people, like, how do you learn? And the answer is like, be present in the

information as it moves, you know? But that, and that's what I felt like I was in, I was at Shop Talk in Vegas when this like all started.

And I like, I honestly started feeling like this, like. No, like the whole world is moving and I'm, I'm not present. And so I think there's this sense by which, just like being amidst it and watching as people solve problems and 'cause the other thing that's gonna help you or helps me to do, I'll say, is that like, is to think about how people approach when they encounter a roadblock.

Like, and it starts to give you a sense by which, what it looks like for people with high [00:40:00] agency to solve problems. Like, 'cause there's two kinds of people that are like, oh, it doesn't create layered files. This thing fucking sucks. I'm out. And then there's people that go like, no, here's what I did and then I did this, and then I found this thing and then I did this.

And then like, and now it does.

Jacob Posel: Yeah.

Taylor Holiday: And you go, okay, that's what I'm looking for. The other thing, as a leader, it's like that's what I want inside of my organization is like,

I don't need the person who tells me it just makes flat files. I need the person that found the way to slice them up. You know, like,

and so I think there's just something about watching those kinds of people that go, oh, if they're around here,

we will be in this problem appropriately.

Jacob Posel: I mean, there's tens of thousands of talented people who are just like dying to show their good work, you know? And like, that's how I kind of was able to like create a little bit of a name for myself. It was like I had no followers and I, like, I was just dying to show my good work and like, and then I was able to, and there's so many people like that, more talented than me doing amazing work and like commenting it and showing people and replying and quoting.

And yeah, it, it is an amazing place to stay close to the action. I actually deleted it from my phone because I [00:41:00] thought it was getting too much, and then I was like, nah, fuck that. I have to stay close to the action. 

Taylor Holiday: It is. '

Jacob Posel: cause I would've missed this.

Taylor Holiday: Yeah. And it like, it's consuming for sure. And you wanna figure out how to make sure that you can pull yourself out of it. Because what I find is that like when moments like this happen, I get to this place where like I could be physically president of a conversation, but my mind is still so like, this stuff is so

strongly compelling in this moment that it is turning it off becomes challenging too, as it's so separate.

Thing, but awesome man. Well, look, I appreciate you. We'll have to keep having you come back in and check in as you keep smashing it and bringing exciting news to the world. And it's gotta feel at least somewhat rewarding. You get to drop a couple, I told you sos that I like to think I get to too.

That like this was the inevitable conclusion of where this was always headed. And it's it's cool to see it come to life.

Jacob Posel: Yeah, I mean, we were, we were saying it, we were saying it. I pulled up the receipts and people were, they were still doubting it and like, yeah, I'm very, very, I'm glad that it actually came because yeah, it's a big, it's a big change to everyone's lives.

Taylor Holiday: Nice [00:42:00] man. Well, you are the the King of Vibe marketing. Congratulations on your your forever marble bust that you created for yourself that says that you've invented it. So I guess you can stake your claim. But thanks for stopping in minute and I'm sure we'll talk again soon. I.

Jacob Posel: Thank you.