Copy.ai CEO Paul Yacoubian on keeping up with GPT’s rapid evolution & AI’s potential impact | E1719

Hey everybody, hey everybody! We all know the AI space is moving at a blistering pace right now, so we decided to do a little ad hoc miniseries on it. We're calling it Innovators in AI. Yeah, very creative. Every week I'm going to interview a founder in the AI space here on This Week in Startups, and every founder that joins is going to be building something awesome. That is the criteria. We don't want just pontification, we want to actually see real products. And today We have the CEO of Copy AI, Paul Yacoubian. We have a great conversation about the insane pace of AI, as I mentioned, and building products for a specific niche. In this case, writing. Copy AI, right? They do writing on top of various LLMs. He explains how he's dealing with the fact that, hey, he started, I think, with GPT-2 or 3. Then they went 3.5, 4, 5 is coming out. Bard is coming out from Google. So many different LLMs out there. How do you actually build a product that people are willing to pay for when these core platforms are moving so fast that they're going to absorb the innovations of people building on top of them almost in real time? This is a major issue for founders who are building in GPT or in AI with OpenAI's products. So it's going to be a great and insightful episode. Please stick with us.

This Week in Startups is brought to you by Mercury, where innovation meets peace of mind. Now more than ever, startups need a safe place to put their cash. Mercury offers a simple way to manage bank risk and protect every dollar with up to $5 million in FDIC insurance and a money market fund. Visit mercury.com to apply in minutes. The Microsoft for Startups Founders Hub helps all founders build a better startup at a lower cost From day one, startups get up to $150,000 in Azure credits, access to OpenAI APIs, free dev tools like GitHub, technical advisory, access to mentors and experts, and so much more. There is no funding requirement and it only takes minutes to join. Sign up today at aka.ms/thisweekinstartups. And Mayfair helps venture-backed companies protect and grow their cash automatically. Diversify where you hold your cash, increase your FDIC insurance coverage, and earn up to 4.35% yield on deposits. Go to getmayfair.com/twist to get started today.

All right everybody, next up on the program is a company that's been doing generative AI, specifically in copy and words for a couple of years. I met Paul a couple years ago when he was raising money for Copy.ai. My bestie David Sachs from Craft Ventures is an investor in the company. And I remember meeting you, Paul, and Paul's last name is Yacoubian, or Yacoubian, right? That's right. That's right. There you go. Yeah. Um, you, I remember, were working on a couple of experiments when we met. One of them was like doing taglines for using— That's right. I don't know if you were using ChatGPT at that point, but when did you first become aware of ChatGPT and/or generative AI for copy?

Good question. So I started using GPT-2 in the end of 2019, and we were using it to create startup ideas. So we were, it was insanely creative, but it was really bad at telling the truth because it wasn't trained on as much data as GPT-3 was. And it was immediately clear that the creative power was where the value driver would come from with the generative AI models. And I thought it'd be 7 years away from being commercially viable because it would just make stuff up and you had to really help constrain the outputs of it to drive value from it. So I was creating science fictional products and just had this explosion of ideas.

What was the interface like for that ChatGPT-2 at the time? Was it all like a dev tool? Because I remember it was a lot of startups were trying to get access to it. Yeah. GPT wasn't really available.

It was actually open source. So you, somebody had to set up the model, host it, and then build a UI for it. And the simplest UI was really just a text box where you would type something in and then just click a button and it would autocomplete it for you. And so I was using that, but I wanted a lot of ideas. And so I kept hitting the button and then deleting what it printed out and then hit it again. And I was like, a search interface would make way more sense, like a UX interface for brainstorming. And so I thought it would be about 7 years before the model would actually be good enough because only about 10% of the results even made sense at that time with GPT-2. And then fast forward till July of 2020, which is only like 6 months later, and they launched GPT-3, which is 100 times the size of GPT-2, and that all the numbers flipped. So all of a sudden, most of the time it was making sense what was coming out.

Ah,

got it.

So when you were doing 2.0, and this is really interesting to look at the history of it, and I'm sure people are going to study this rapid pace that this has gone on. You, there was no web-based service, there was no cloud-based service. You had to download it and then Did you have to provide a model or was there already some training in it based on some model?

Yeah, it, they had pre-trained it and they had open sourced the model. So somebody else ended up hosting it and providing that interface.

Got it.

And so that's what I was using at that time.

Were they upfront about what was in the model or did they just say, hey, we trained it on some stuff? Um, or were they explicit? Hey, we downloaded the Wikipedia. Hey, we did Quora. Hey, we did Reddit. Were they at that time? 'Cause that's something that I've been trying to get my head around and hopefully you can educate us in the audience as to What exactly, how do you know what's been trained on?

Yeah, they, I think in the paper listed out a number of the sources. One is a Common Crawl dataset. It's just a huge, huge file that's open source. You can download it and you can use that to train the model. Wikipedia was a big component of it.

Common Crawl as in a common crawl of the open web. So just

a bunch

of web pages that are in an open source search engine of sorts.

That's right. They got Reddit and I don't think they got Quora at that time. I think—

Yeah, probably had a paywall or something, or a registration wall so they couldn't get in there.

Right. I think since then they have, they have probably gotten access somehow to Quora's dataset.

So you see 3.0 and it goes from 1 in 10 times to getting it right to 1 in 10 times getting it wrong. Yeah. Or said another way, 9 in 10 times at least making sense. Like it wasn't completely Gobbledygook gibberish.

Exactly. Yeah. So the first thing that came to my mind, I was like, this is the next big wave in tech because you have all these new use cases that are unlocked. And I thought we would see immediately, I've tweeted this out, I thought it was a phase change of the internet. So this is the first time the internet can be trained and synthesized into a model and then talk back to you directly. Right? So that's a profound change in the amount of information that you have access to at any moment in time. And given that you're kind of able to talk to the internet, I thought that conversational interfaces would be the most relevant way to access that information that's inside the models. So we built a Slack bot and we asked OpenAI if we could launch it, and their safety team said, no, you can't launch that, it's too dangerous.

Hmm.

And so we had to go back to the drawing board. And we built an app that could just summarize whatever you threw at it. So we launched that, we got a mention in the Wall Street Journal for that. And then we realized, you know, people, that's like a sometimes use case. I mean, that is a component of a skill that the AI models have, but that's not something that people will subscribe for and pay us money every month.

Listen, as a founder, ensuring your cash is safe is priority Number one, it's been a bit crazy out there. I don't want to put caps locks on right now, but I will if you need me to. The FDIC $250,000 limit is just not enough for most businesses. So let me tell you about Mercury. Through its partner banks and sweep network, Mercury customers can access up to $5 million— $5 million in FDIC insurance. That's 20 times the per-bank limit. Sounds a lot safer, doesn't it? Well, these sweep networks protect your deposits by spreading them across multiple banks, which limits the risk of any single point of failure. And with Mercury Vault, any funds above the FDIC limit can easily be invested in a money market fund, mostly comprised of US government-backed securities. So it's easy to get started with opening an account. You can apply in minutes, and many customers are approved and onboarded in less than 2 hours. Mercury also offers great resources for founders, including my favorite, Mercury Raise. This program connects founders with investors to help them raise funds. Head to mercury.com to join more than 100,000 startups that trust Mercury for banking. And if you're interested in Mercury Raise, applications for both seed and Series A are open through April 20th, right? So you got to get it in there by April 20th. Mercury is a financial technology company, not a bank. Banking services are provided by Choice Financial Group and Evolve Bank and Trust, members FDIC. When it's summarizing something, explain to a layperson what it's actually doing. So you have a paragraph, let's say you have a, I don't know, a movie review from the New York Times. It's 1,000 words about some, you know, really detailed movie like Tár, right? Academy Award, uh, Oscar-nominated Best Picture. What does it What is the— what is the, um, what is the model actually doing when it is summarizing?

So there, there are different techniques for summarization. For these models, um, you know, based on neural networks, the neural net is learning skills. So it is— it has learned skills that you'd have otherwise needed specialty natural language processing algorithms to do. And that's really what sets it apart. So when you had 176 billion parameters in the model and they're structured into layers, these layers get really good at doing certain things. And for the summarization tasks, it can learn what summarization means. Like, what does the word mean to summarize something? It can pick that up from the corpus that it's trained on. So all the text that it's trained on, and then that is a task it can do It can do pretty well, which to me, I was like, okay, now this one model we can use to do all of these natural language processing tasks without needing a bunch of NLP engineers on the team. It's literally one API call and I can just prompt it in natural language and tell it what I want and it'll give that back to me. So the first, that was a major unlock because it suddenly made the process of building an MVP a software MVP that uses natural language technology, made that really easy. So you could build a prototype and launch it. And so when we built our next prototype, it was called taglines.ai. It was just a tagline generator that took 48 hours from beginning to end to build the entire thing. And we launched it 48 hours after saying, okay, maybe people would buy you know, maybe marketers would buy words if we can generate them. And so within the first week, we had validated that not only would you have professional copywriters using these tools, but you'd have marketers and freelancers and small business owners, students. So like, that was a huge validation of the market size and the TAM potential because it was such a wide user base. And then we also validated that people would use a tagline generator to do like email subject lines, Headers for websites, just all kinds of like PowerPoint slide, you know, titles, and really validated that the number of use cases were really wide for that as well. And then threw up like a Stripe paywall and then validated people would sign up for a subscription. So we started charging $3 a month. We got some subscribers, raise it to $6 and then raise it to $10. And we continued to get subscribers coming in. So at that point we validated one, that it was commercially viable, that you had a wide market and a wide set of use cases.

Right. And that makes sense. There are copy editors in the world. You discovered this is working 9 outta 10 times, and if you can make them more efficient for $10 a month, if they, the average copywriter, I think freelance probably gets paid $50 an hour, $40 an hour, $60 an hour, something in that range. I mean, it's not even 1 hour of their time. So that's always one of those great tools for founders is to say, what does this person get paid an hour? Okay, this attorney gets paid $1,000 an hour. Save 'em 1 hour a month, they'll pay you $1,000 or they'll be okay with it. Right?

Exactly. So when, when we went through that exercise, it was really clear that from a, all the way from a student to a professional would find value in these AI models. And then the thing that we learned from going from 2 to 3 was that if you throw more data at it, it's gonna get even better and better at performing these tasks at an expert level or beyond.

So now when we get to GPT-4, to get back to this sort of these jumps, there was 3, I think there was 3.5 and 4 in terms of major releases. You are working directly with the OpenAI team. You have some insider access to it, I understand. So you were kind of in, in touch with them and, and seeing what was coming. What was coming between 3 and 3.5 and 3.5 and 4?

Yeah. So for 3.5, they ended up tuning it to human feedback. And so what that would do is it would allow you to kind of describe the intent that you had, and you wouldn't have to prompt it with examples as much. So when we first built our tools, each tool, you had to give it examples of what good looks like, and you had to make sure that even the examples are very diverse. When they ended up fine-tuning it on human feedback, they kind of program that in across a wide variety of use cases. And so that enabled the ChatGPT experience where you can just tell it what you want and it'll figure out and actually create that content that you want.

And so when you saw 3, 3.5, 4, maybe you could describe for the audience. Yeah. The step functions there. You said before, hey, you know, it's right 1 out of 10 times, then it's right 9 out of 10 times. Now let's go to 3.5 and 4.

3.5, yeah, was awesome. It was like a big leap, and a lot of the complexity in prompting it kind of went away. A lot of even the use cases for more fine-tuning of models. So you'll see a lot of startups talk about how valuable their fine-tuned models are, like, oh, we have proprietary models. Well, at the end of the day, there aren't that many models that are Really relevant for, for you to have fine-tuned.

Explain what a model is again to a layperson and what's an incredible example of that where you're restricting it and building a model that maybe results in better output.

Yeah, so the, these are foundational models, so these are built to be general purpose. That, that means they can do a lot of things at, you know, a pretty, pretty solid rate. What they figured out was that the general purpose model could actually outperform specialized models if they kind of made them bigger and trained it on more data. And that is nuts. That's nuts. So I don't know if you've seen some of the GPT-4 metrics, but they've said, oh, well, it passed the bar exam. It passed like these, all these exams, right? It's getting better across every single category.

So let's pause on that for a second.

Yeah.

You used to build in the industry what was called narrow AI.

Yes.

Narrow learning. So, hey, we're gonna learn how to beat, um, you know, humans at poker, chess, Go, you know, pick a video game, Fortnite, and we'll really just work in that narrow subject area. Now, this more general model, um, we'll A little confusing to use the word general because it's loaded, because there's general AI as a concept, which is thinking like a human, but their average or their default model maybe is a better word for it. Their default model is better at doing the LSAT or the SATs or the bar exam than a model that was trained just to do the bar exam. That's what we're— they're learning with Four.

That's right.

Yeah. And you seem to say, hey, that's insane. Why is that insane? Why is that insane?

Uh, it really does level the playing field if you have access to those, the GPT-4 model. So you don't need a whole team of AI engineers. You don't need a whole team of machine learning engineers. You can actually build, you can just go straight into building the application. And it's—

Yeah, keep going. I

was going to say something that I think larger companies are running into issues around is they've kind of delegated the, like, hey, AI team, go figure out this generative AI stuff. But the AI team, you know, if the AI team comes back and says, you know what, we can't beat GPT-3.5, we can't beat GPT-4, they'll get fired, right?

Right. That's fascinating. So if I was working inside of I don't know, Amazon, I'll just say Walmart. And I said, hey, our internal team—

Amazon's a great example. Amazon's a great example.

Yeah. I'm in Amazon. I want to optimize Amazon Prime members to show them better reviews and show them better Q&A. So I take the Q&A section on an Amazon page and the review section, and we'll summarize it and make it better for users. If GPT-4 can do that better than the internal team, Why is there even an internal team? Yeah, that's pretty wild.

Depends on the use case. So if you looked at like the iPhone voicemail system, you know, try to transcribe a voicemail message, they had, I don't know how many thousand AI engineers at Apple, and then OpenAI had a team of, I think, 2 or 3 people that figured out how to complete the voice-to-text task, and they ended up launching a model called Whisper, and they open-sourced it. They literally gave away the technology, and it dramatically outperformed all of the existing technologies, all of it.

All right, everybody, our friends from Microsoft are here. Tom Davis, a senior director at Microsoft for Startups, and you're a former founder. You are here today to talk to us about The giant leaps that Microsoft has made in the AI space. What does this mean for startups? I see a ton of different tools. I've been playing with ChatGPT-4. I have a paid account, but I'm also seeing things happen with GitHub.

Absolutely. So the work that we've been doing with OpenAI over the last few years has really set ourselves up with a foundation around— we've built this sort of AI supercomputer from the ground up. And we've been looking at everything from GPU configurations to networking and things. And really what we're now able to do is sort of allow startups to access all of this innovation through our Founders Hub. So we've been building this to do something at scale so startups can now build their own AI applications and, and build out and train LLMs as well. And this has really helped us to become a far better cloud for AI broadly, and being able to drive that down to the startup ecosystem is fantastic.

It's open to everybody. There's no funding requirement. You don't have to be anointed by a VC. 5 minutes to sign up and you get 6 figures of benefits, Azure credits, GitHub, OpenAI APIs, which everybody's really having fun playing with, and so much more. So go ahead and sign up right now, aka.ms/thisweekinstartups, aka.ms/thisweekinstartups. Thanks so much, Tom.

Thank you.

So as we're looking at this, yeah, this is almost like the snake is eating its own tail kind of situation. We're all going, you know, uh, copywriters, you name your company's CopyAI, you got $12 million in recurring revenue, uh, and your premise, which turned out to be absolutely correct, is, hey, this thing is advancing fast. Copywriters are going to use this tool. They'll be willing to pay for it. It's going to make 'em bionic and make 'em superhuman. Um, won't replace them, but might replace the bottom third that are terrible at their jobs or make the bottom third great at their jobs or good at their jobs. Was that your central tenet, um, about this? Because the replacing of a job or elimination of the job, in other words, I'm the developer or I'm the CEO, I'll just ask ChatGPT for copy and I don't need a copy editor. Uh, what, what is your belief now? Let's say copy editing, your wheelhouse. Is copy editing gonna go away and the CEO or the sales team just does the cop— asks ChatGPT to do it? Or will there be bionic copywriters who use this as a starting point?

I think before, you know, before we launched CopyAI, I had no idea what copywriters even did.

Hmm.

So it was, it was never about copywriting. It was, that was basically the first use case that I thought would be relatively translatable against the models. But the thing that really motivates me and my co-founder is the idea that people have great, like a high degree of creativity, especially kids. Yeah, they're super creative. I've got 2 daughters that are extremely creative. And then they, we pre-train them through school, right? Mm-hmm. Just like these models are now pre-trained on the internet. And all that pre-training kind of stamps out the creativity. And by the time they graduate, they're very tracked in the things that they want to do. So it's like, oh, I'm going to be an attorney, or I'm going to be an accountant, or I'm going to be a software engineer. And I think that that whole process is going to come to an end here because GPT-4 can do all of the liberal arts. Better than any expert could. Right. And now that's a tool that's in your pocket, which means we don't need to train people on a lot of these very technical tasks that we used to because we've now trained computers to do that repeatedly. So there will be a very challenging transition for a lot of industries, a lot of institutions like higher education, education in general. And we're gonna have to figure out, okay, what's the best use of a person's time? Well, that usually comes down to the individual. And what I'm seeing, you know, my whole hope here is that people can be way more creative because each time they are creative, they have an outlet to actually build something. They can actually get things done.

So a non-copywriter, a non-writer, somebody who's just not good at words, can now go in there and be good at words. So you take the—

Exactly.

80% of people who are not good at that, now they can be good at it. I may not be good at drawing or illustrating, but I can pop up Stable Diffusion or any number of tools and I could be a good illustrator. So now you've got the 99% of people who have no illustration skill, can't draw for not, and they can be creative. So it's not that everybody's job goes away, it's that everybody gets good at everybody's job.

Yes.

That's an interesting way to look at it.

Right. And that's incredibly powerful, especially when you're looking at entrepreneurship. So there, if you've ever tried to create your own website, for example, that takes a lot of time and it can be very stressful for somebody that's trying to get something off the ground. That is now going to be pretty much instantaneous. Like, oh, I need a website. Boom. I've got one. I need a presentation. Boom, you know, I can just generate a whole, like my whole deck. Um, and so—

And then work from there. So you're starting with a base. Yeah. Yeah.

And so—

Which happens now, you can go on Squarespace, pick a template, and now you'd be able to, a future version of Squarespace will be, hey, make me a website and show me 5 different designs. I need one that's funkier. I need one that's a little more avant-garde. Um, now the challenge.

The challenge is what kind of business are you going to create that's going to actually be able to compete against what the AI models can do?

Yeah. Well, and to your point, the AI developers who are making this are literally working themselves out of a job. The better they do, the less they're needed because the general model can do better than the specific model. So at least in that case, the narrow specific models are being subsumed into the general models.

Yes. Yeah, you're seeing that. And then you're also seeing it in the software itself. So I'll give you kind of a run-through here. So software is— why is software valuable, Jason?

Because it makes people more efficient, which means it reduces the cost of any good or service.

Right. And what, like, why software is valuable is because it's reusable. So your investment to build it—

That's a big part of it, yeah.

Right, is the reusability. So we're now heading into a world where software can be generated on the fly, right? And it's almost the cost of creating new software is going to zero, right? That means that the returns on investing in that software should also go to zero, which causes some problems in the venture capital community. I'm an angel investor.

Yeah.

You know, you're an angel investor. So we really, this whole thing is going to get transformed from a technology standpoint.

Fantastic. I mean, if you can build a company with 2 people instead of 10 or 10 people instead of 30, well, that means you need less investment and the earliest investors are the only ones who get to a seat at the table. I'm fine with that. I'm fine with people who do Series Bs and Cs getting Blocked from having an opportunity to invest because the company's too profitable. Is that sort of your premise? It takes less money to run this, just like cloud computing dropped the cost of it.

Yeah. So one, it should expand the number of people that can build things, which should be good overall for entrepreneurship. 2, like you said, on the growth stages, that's going to increasingly make, I think, a little bit less sense. And you've seen, you saw YC shut down the continuity fund. which is their growth stage fund this week. And so in that, and then when we look at the AI landscape, a lot of the repeatable tasks are in the sales and marketing space. So like personalization now can be done on a one-to-one basis at scale, even things like querying databases. That's something that people had to do. On their own. So you'd have to figure out, well, how do I write the right search query to get the right results, maybe on LinkedIn? Um, so what, what you're going to see happen, and this, uh, you know, chat is a great example of this. We'll express intent, right? Which is what our query is when we type into chat. And then the AI will like figure out what it needs to go do to get. Mm-hmm. us back what we're looking for. And so that, that is a world of AI agents where we're letting the AI go and do things for us. And that world looks very different as well. So AI doesn't need an interface to go grab data, right? It can just query APIs directly and go get the information it needs.

If it's allowed to, if it has access to those, right?

Exactly. So I had a tweet about this. I said the biggest By the end of the year is my prediction, the biggest search engine on the planet won't be Google, it'll be Bing. And it won't be because people are using it, it's because the bots are using it.

So, so many bots are doing queries that it's just going to increase the size of the corpus and the learning model.

The corpus kind of grows when you publish things. And so there are going to be agent flows and bot flows that go and create more content and new content. But for the, on the retrieval side, that a lot of those queries are going to get sucked into, you know, powering some result, right?

And then that result gets reindexed. It gets reindexed. And now people are speculating it's going to pollute the search indexes to the point of absurdity. Somebody like yourself or some group of people offshore, just like they created content farms, they could literally just start creating They could create an AI right now with ChatGPT that registers 100 new domains a day, makes 100 different recipes and then articles about those recipes, and then does the same thing every day until there are so many recipe sites that the internet is flooded with garbage and nobody knows which one actually works. That's the fear. That's the negative thing.

Yeah. Google kind of knows.

Google knows. How do they know?

Well, they have They have the original index. They have, yeah. And they, they can detect AI-generated content overall.

How do they do that? How do they generate, how do they know it's ChatGPT-generated?

There are signatures in the outputs that, that allow you to calculate. It's a perplexity score sometimes. And that's, that's an actual measurement of how likely the next word is, is to follow the previous one.

Got it. So it's too perfect.

It can be too perfect. And you can, you can like make it not be as perfect just by giving it certain instructions and say, hey, mix it up a little bit.

I think this is how they, they caught the chess cheater kid was he was doing the book move.

Yes, he was. He was 99% accurate.

Yeah. How often does somebody do the book move? And I play in the Chess.com thing and it will tell you like, this is the book move. This is the, You know, game theory optimal move, basically. And if everybody's doing game theory optimal, so now you have to put something in there to randomize to not do game theory optimal. So maybe at some point it figures that out. You know what's interesting though? They said when the digital camera came out, uh, that this would change Hollywood forever, that digital filmmakers— and Wayne Wang did a digital film, um, and Bennett Miller had done a digital film on— this is all on the VX 1000, I believe, was the name of the Sony camera.

The Sony one, yeah.

It became super popular in the late '90s in New York and all these—

And skateboarding.

And skateboarding, yes. Well, skateboarders use it, right? Yeah. And it was like, this thing is so good that when you blow it up and you put it on a projector, you can't kind of tell, or most audiences can't tell. A cinematographer obviously could. And then there were filters in, you know, different pieces of software that would make it look even better. Anyway, the point was, because you don't have to worry about film stock, developing it. The cost just got sucked out of this. You could record forever. Even a bad actor could just keep take, doing different takes. You would get there and it didn't actually change the world. I mean, it did create YouTube, it did create a lot of other content. So maybe in that extent it did, but it didn't change film. So what's different about that argument than your argument for, um, you know, AI replacing everything?

Well, the, the iPhone won, right? iPhone won the camera world. It destroyed it. It also won music. So my base case is that Apple wins probably this too, to a large degree. One, because the phones are powerful enough and the models are going to shrink small enough that they can fit on a phone. And so even some of these image generation models, they can fit now on the phone. Text gen, same thing. On a long enough timeframe, they will fit on the phone. And I think that, you know, that's always crazy, but then people just get used to it and they're like, oh yeah, iPhone 20 will have whatever it is native built in.

Well, just like laptops now have Wi-Fi built in. There was a time you had to add Bluetooth with a dongle or Wi-Fi with a dongle or a PCMCIA card, whatever that was called. You know, you'd have different card slots on your laptop to put in different peripherals. Now it's all on a chip somewhere and You don't have to do anything for it.

Exactly. I have one more point here. So my kids are 8 and 5, and they just think computers can generate images. It's just totally natural to them. You know, and in the, I think it's adults and people that have struggled to use computers their adult lives, it's like, this is, these are the people that have, you a challenge, like accepting that AI is going to power computers to do these tasks for us. But kids are like, yeah, of course, of course. Yeah,

of course it's going to. Yeah. I mean, they—

It's just like,

well,

it's so obvious that a computer could do it to them that it's not even special or interesting.

You know, if you are of the age where you remember the internet happening and broadband, and there was a moment in time where I was on a college campus in the late '80s and I had seen higher speed internet speeds, BitNet and ARPANET, and people only had dial-ups. So I was using dial-up at home, but I was like, you know, this, this can get a lot faster. There are faster systems than this. And I had installed Token Ring and Banyan Vines and then eventually Ethernet in local area networks. And I'm like, you know, this thing's going to be able to carry images and pictures and sound and eventually video. And so it's kind of the same thing. Like you, your limitation right now That you're like, oh, it can produce interesting copy that's sometimes good, sometimes bad, okay images, and once in a while a 3-second looped video. I mean, you're going to be able to talk to this thing and just make your own Star Wars story and it'll be indistinguishable than, I don't know, the original series or whatever, 10 years ago series without any polish on it. I mean, that basically is— doesn't take much of a jump actually. Only takes a huge jump if you haven't watched rapid

Right. Well, I think in the next year you'll see it, um, probably just take over business in general. So we know it can generate business ideas. We know it can create websites from an idea. We know it can figure out who the core persona is for that service.

Hmm.

We know it can access APIs that would retrieve, right, who the target contacts are. We know it can write personalized emails to them, right?

So it's going to come up with an idea, you make your SaaS startup, and then it starts selling it for you.

It's

all in the same sequence of prompts.

Yeah, that's all it is. So that's all that is here today. It's just a matter of infrastructure, and that's something we're working on right now.

So what I mean, What you've described means your own business becomes commodified. So are you saying that your own business is going to, or are you saying your business will be commodified every 24 months and you'll just have to extend it?

No, I'd say like most businesses are not commoditized by that loop. Most businesses have some kind of like real-world thing happening. Let's say like JP Morgan, right? That's not, you can't really replicate JP Morgan. With that kind of system. But there are a lot of, one, services that can be replicated, like specialists, consulting. A lot of that stuff can be replicated. A lot of software companies can be replicated.

Yeah. I mean, you're talking about regulation. I mean, you could, I mean, somebody's going to put AI and Web3 together and crypto and be like, make me a crypto that does X, Y, and Z.

They're trying to, I mean, they're trying to shut all that down with

Regulation.

Regulation.

The, the crypto choke point. Yes, if you fuck around too much, you will find out what the government is capable of. When you're a founder, priority number one is hitting your payroll. This means you need fast access to cash no matter what goes down. So startups need to hold their money in more than just one FDIC-insured account. We all know that, right? Been pretty choppy out there of late. So let me tell you about Mayfair. Mayfair is a cash vault powered by Stripe, and it's built for startups looking to diversify and grow their funds automatically. Takes 10 minutes to set up. It works alongside your existing bank accounts and, most importantly, FDIC insured. Here's how it works. You sign up for a new Mayfair account. You keep operating cash at your existing bank and deposit the rest at Mayfair for safekeeping. Then Move your money back whenever you need it, or let Mayfair manage your balances automatically. Mayfair can help you earn yield with no investment risk or lockups. Earn 4.35% APY on daily balances paid for by the bank. Your cash stays in cash and is never invested into money market funds, bonds, or other instruments. Withdraw your funds at any time, just like a checking account. Mayfair is designed to keep your money safe. They partner Stripe to move your money securely, and your funds are held in the custody of Evolve Bank and Trust, which powers companies like Mercury and Wise. Your funds receive automatic pass-through FDIC insurance. So here's your call to action: Twist listeners that sign up today earn 4.35% yield on their daily balances. That's more than 10 times the national average savings account. Go to getmayfair.com/twist to start managing your cash like a Fortune 500 company today. That's getmayfair.com/twist. Okay, so show me CopyAI and what it's doing today, because you and I were talking at South by, uh, and I was like, huh, a lot of the stuff you showed me is now built into ChatGPT-4. What does that do for your business?

Yeah.

So I'll give you that same question, which is, hey, you know, the stuff you were talking about 2 years ago, to your credit, I literally put everybody in my company, I said, everybody has homework this weekend. We're a finance company. We all work 6, 7 days a week. If you don't like it, don't work at a finance company. Um, and I said, just spend an hour or 2 playing with, uh, ChatGPT and, um, tell me on Monday how it helps you do your job faster, better, or—

What did they come back with?

Uh, the sales team started doing prospecting in it.

Mm-hmm.

And other people started doing marketing plans in it. And I haven't gone through everybody's responses, but, uh, the, my team on this podcast has been doing podcast notes with it and that works really well.

Mm-hmm.

Which we thought was a goof 60 days ago when ChatGPT 3.5 came out. I was like, just ask you what the revenue of this, you know, of Netflix was the last 5 years. And it was wrong. And then ChatGPT 4 comes out and it's like, okay, it's right. So it's starting to, things that we even noticed 3 months ago, 4 months ago were wrong. Right. So you could, and then we asked it to take, you know, um, we will take a Zoom recording of a meeting with a founder, transcribe it. And this past week we took the transcripts. Which now Zoom is doing. And then we had ChatGPT-4 summarize it and summarize it in 3 different lengths. So now in our Notion instance, not only do we have the video with the interview with the founder recorded, we have the transcript and then I guess 2 or 3 versions of it. So I could read the short version of it and I was like, it's pretty accurate. I mean, I wouldn't rely on it to make an investment decision, but certainly would be interesting to read before you meet with the founder.

What about GPT-5?

It can do it. You have to assume GPT-5 will be extremely less error-prone and you will just trust it. Just like GPS. The first GPS, remember people ran off the side of roads? It'd be like, turn left and somebody would wind up in a cornfield and be like, it said turn left. And

it's like,

hey dummy, you still have to look. Just also like autopilot. It's like, it's going to stop at this red light unless like there's a Christmas tree at the red light, you know, confusing it. Um, but it feels like you can trust it. I think trust is going to be the main issue is, can I trust this by default?

Yes.

Do I trust it more than a human? And I trust self-driving on a highway more than I trust a human today. That's, you know, yeah.

And, and we're, we went through the, that valley where the quality was lower than people can produce. And that was like the GPT-3 valley. And that's where a lot of people are saying, oh, well, the content's terrible and like it's awful quality and it's just going to pollute the internet. That's a very short duration until you basically can replicate all the steps that a person would go through to write a high-quality article. You just replicate that at scale. And I think one of the key points that even ChatGPT-3 is not is kind of missing right now is the scalability component. So in order for you to get value out of it, you have to manually query it. So your marketing team has to say, okay, we're going to go do this, or your sales team is manually configuring it. The area of opportunity for, in our kind of viewpoint of the future, is really scaled, fully automated workflows. So you give the AI You kind of tell it what you want to do, or, or over time, just give it access to your systems, and then it will go and configure itself, and it will try to improve your business for you autonomously.

Hmm.

So you can't really get there from chat. And so, but part of the backend infrastructure is very similar because in the world of autonomous business process, agents, that is like an agent building platform. And so you do need to say, okay, I want to do this. Here's, here's the, here's the list of tools you have access to. Go do it. Now, instead of reporting back to me, report back to yourself and then figure out how you can improve it from there. Right. So what you want to do is, is build these, um, you know, extend the data flows that you have coming off of your current software. So you have a newsletter, right?

Yep.

What do you do with the information, the data that comes out of the newsletter performance?

Nothing. Don't care.

Right. So is there value there?

Sure.

Right. And it's, and it usually comes down to like the team just doesn't have enough time to do everything.

Right. It's an afterthought to think about and to coach yourself. There's no person on there. So you could say, every time the newsletter comes out at the end of the week, tell us which stories got the most clicks. And then what are stories like those that we can do more of? Boom. And that's just set every week and it just sends them an agent. So it's like Google Alerts or Magic Leap had this idea that you'd have agents out there doing things for you. But if you take a Google Alert, which is on your last name or your company name, There's a Google alert that can do more advanced things.

Yeah, exactly. And, and even the newsletter, maybe you have like very high value contacts in your, in your newsletter database. Sure.

Yeah.

Right. And you're like, hey, I actually want to personalize theirs. So I want to take this content and just literally personalize it down like one-to-one to them because I want every single message they get from us to be extremely high quality.

Ah, interesting.

Right. So those are systems that are going to be, you're going to be able to build. We're building out, um, we have pilot enterprise partners that we're working with now that are deploying some of these workflows into their existing business processes.

Fascinating.

The, the thing that compounds is once we set up a workflow, um, let's, let's call it like the newsletter use case. Most newsletters would need it. And so that is the thing that we're gonna copy.

That'd be a template in your library. Like Notion has templates or any other platform has templates. And yeah, that gets super interesting in a creative. We did, uh, just as an, an example, here's a couple of, you know, when we do, when the, when our researchers are gonna do a Google search, like I'm doing a live show and they're like, hey, when did Netflix launch? When did it pass 100 million subs? When did Disney Plus launch? When did it land, do 100 subs? It's now getting that right. It wasn't getting that right previously, but here's examples of ChatGPT doing that. And it does seem like Bard from Google, which I've been playing with, has more recent information. So in your mind, how far behind is Bard, do you think, in reality?

Probably the measure that matters is the user base for chat programs, I think. That would, that would be the, the difference maker there.

That ultimately will determine who wins. So then Google does win because even if OpenAI has tens of millions of users, Google has billions across Chrome, Android, YouTube, et cetera. So Google wins.

It's, I wouldn't count anybody out. I wouldn't count Google out. I mean, Google has, they have all the datasets that matter.

Yep. So they should be able to go faster and they have the users and they have the browsers.

They have to really get their vision together about what they're building. And for them, their goal was to make all human knowledge accessible and useful. That was their mission statement. And search does that, right? It does that along one dimension, which is You can search for something. So it's like, I'm going to pull information out and they show you the list of links, which is great. However, the idea that it's making it useful, I think they've— they're failing to deliver on that function.

The output that you get back.

That, and then also just think about how much value is sitting on the internet that you don't know about that would create value for you. Right. And until now, we've never had a way, like a technology that could take all of that information and directly apply it and make your thing more valuable, whatever it is that you've built and have been working on. Right. So even for you, you know, you're the media company, content company, you're trying to read a lot of stuff and figure out what's important, what's valuable. And there, and I'm sure in the past, Let's call it 4 weeks, you see more launches than you can even keep up with.

It's crazy. I mean, the number of people launching verticalized AI, like if I had a nickel for every company that's going to respond to sales emails or customer support emails in a better fashion, I mean, it's in the hundreds now. Yeah,

right. And the thing we ran into, we were trying to build verticalized solutions and we realized that it's It's just basically the same backend infrastructure that you need to do any of them. And so rather than a company, a large, let's say, bank trying to evaluate 3,000 verticalized solution providers, they'll end up opting just for a platform. And you've seen this happen over and over again. So Segment was a CDP layer, customer data layer, that they didn't have to build. Hmm. Snowflake was a huge, like, data, you know, data warehouse in the cloud. They didn't have to build that. ServiceNow, one of the originals, right? Just scaled so hard and they were very use case agnostic. And that's, that's very appealing for large companies because they're always trying to reduce the number of tools that they have at the company level.

Yeah, for sure.

So that's,

I mean, Microsoft Office, perfect example. You have bundling of stuff. So show us your product and where you're at with it. And what people are using the most. And it, and specifically how does it differ from ChatGPT-4? Because that, as I said, my premise was, or my question to you was, was like, does this thing eventually collide or are you just in an arms race where they're building this sort of web-based service and then they need people like you to build stuff on top of it?

Yeah, it, it works both ways. I think there's plenty of room in, in this you know, in these categories. This is our chat product. We've connected it to the internet. Let's ask it. So this, you know, SVB got sold today.

Yep.

And so this would be, you know, a real-time query that we can run on it. You know, most people were using like our— I'll show you these tools here. Second. So over time, all the latency drops and the costs drop. And so for the, uh, for this result, you know, you're getting the actual articles, right? And—

Give me the citations. You're surfing the open web, so you have Google News or some news feeds fed into here, or using ChatGPT-4 is like, hey, go check the web.

We can do anything. We could do any of it we want. So you can hit, like, you can do it in real time. You can do like a search API that you can add, add to the backend. You can use the Bing API. Um, that stuff doesn't really matter too much. Like you can do any, anything you want and then you can bring all those results back in. And for, for this one, you know, it answered the question really fast.

Perfect, right? Yep.

Right. And when GPT-3 launched, it would make something up. It would just literally make up like the wrong answer and be very confident in it.

Yeah.

Right.

What's the word for that? Is that what they mean when they say it's delusional or it's, uh—

Hallucination.

Hallucination. Yeah. It thinks it's right. It has no idea. Yeah. Okay. So this is impressive. What next?

So what's happening in the backend is actually where the secret sauce gets built. So when you, you know, when a user comes into chat and they ask a question, the, you know, in the Google world, it would just hit up its index. So the index would say, okay, what's the closest set of results that I should show to this user based on that query? In the generative world, you can build these agents so you can go try to solve the problem. And the more tools you give that agent, the AI agent, the better it can solve the problem. So it is choosing now what route to take. So these are not pre-programmed steps that you build into it. So

it's saying, I could go to Wikipedia, I could go to Google News or Bing News, or I could search the open web. I could do a web crawl. I could do, excuse me, any number of things. I searched Twitter. Search Reddit or Quora, and it somehow figures out, you know what, this news service is better. Or is it blending the news service with the web search?

It's going to do a little bit. It can do whatever it wants.

Yeah.

So this is fundamentally a new way of building backends to software. The other thing that, you know, ours can do is it can recommend new steps. So it can add, you know, we can figure out what tools it wants and we can give it access to more tools. So in some cases, people wanted access to like LinkedIn data, right? So we can go get a link, an API that queries LinkedIn profiles and returns that into like a JSON format, which is like a data format. And then now we can take that information and do something with it that's relevant to that user.

This is going to make these large datasets invaluable. And if you're going to use them, you're going to need to get a licensing fee from LinkedIn, Reddit, Quora. Maybe you could talk about what's going on in the back channel, because you're probably facing this yourself with those discussions. I understand there's like multiple lawsuits about to drop on OpenAI for using various platforms' data without explicit permission. And once you start charging for stuff and you take $10 billion from Microsoft, you can no longer be like, oh, it's an academic nonprofit. Even nonprofits, like, I don't know if you saw the Wayback Machine, the Internet Archive lost its copyright case, and that's a nonprofit. So this is clearly unfair use. Now, I don't know if it breaks the fair use. We'll have like a, there'll be a big debate coming up, but what's the back channel right now about who gets access to Quora, Reddit, Twitter, pick a dataset, and how will that cash out in your mind? What's the fair route there?

That's a good question. I think over time, you know, the li— I mean, licensing tends to get figured out over time.

Yes.

Music industry, they figured it out. Even in the image generation side now, I was at South by Southwest and at the Shutterstock booth and they had an AI-generated image product. And even them, even though you're using AI to generate it, they would find the images that went into the generation and they will pay a royalty to the Copyright holder of that, of those images.

Yeah. Getty Images is suing Stable Diffusion right now. Yeah. A bunch of open source developers are suing Microsoft's GitHub Copilot. So these are dropping—

Those are, to me, those were like very, they should get sued at least.

Explain your thinking.

For those, the code base, I think I would at least want to see like how those arguments play out. So if you open source your code, someone grabs it, trains a model, and it is replicating your code verbatim, and even in some cases, like the script, the documentation script that was personalized to you, I mean, that's very traceable, right?

Yeah. There's no reason not to put citations in it. Why wouldn't you?

Right. So that same thing happened on the initial Stable Diffusion dataset. So they did You know, they grabbed images and you could see like Getty and the watermark.

The watermark. That was classic.

So we, we kind of avoided doing image gen. I know other platforms wanted to launch that as products, but if, you know, any of these image generators, they'll produce copyrighted content. And the way the copyright laws work is if you are the one that generated it on your servers and you're selling it, Like, that's a— you're gonna get dinged for that.

Yeah, I mean, that's— it's a derivative product that you created. Now, if I were to load, uh, just like on my browser, I can take my browser, I can save a web page on my desktop, that's fair use. Um, and I can even edit the web page on my desktop. It's when I choose to put it out in commerce that your copyright gets, uh, you know, has problems. But if I were to If I were to download the original ChatGPT open source and it's no longer open source, is this like ChatGPT-4 is not open source now? When did it stop being open source?

GPT-3 was not open source. So 3.5, which powered ChatGPT, not open source. 4, not open source. And it's unlikely that OpenAI will open source.

Is there, what's the number one competing product and how far behind are they? Because it would, it seems the world needs an open source version of this. And if I would have run the open source version on my cluster of servers or, you know, my desktop, my phone, and pointed at a dataset that I had access to, well, that's for my own personal use. Uh, nobody can stop me.

Right. You can do that. And even the Stable Diffusion models, they've, they've shrunk 'em down. They can fit on your M1 MacBook or your phone. So no, you know, no kind of copyright issue there. But if you try to monetize that, You do run into it. In terms of open source, OpenAI is still state of the art for their core foundational models. The advantage actually is not on the training side, it's the computational intensity side. So it's how they optimize their hardware, how they optimize the training of the models. Oh. That's where you build a really compounding competitive advantage. Because the bottleneck across the world right now is GPU capacity and availability. So Nvidia has backorders like crazy for their H100 GPU and no one can get it now.

Explain what the H100 is to folks who don't know.

It's a really expensive GPU, so like a graphical processing unit, and it's what powers video games. It does a lot of Kind of matrix math. And that's what these neural nets really need at scale. And anytime a big company wants to launch an AI-powered product, that's just more compute that's needed to serve that up. And I think most people will have run into outages. ChatGPT goes down or OpenAI goes down. It's a function of the limitations of They're just not having enough of these GPUs in their server farm and getting overloaded.

But isn't the Tensor stuff all open-source hardware as well? So other people should be able to build these and compete over time?

No. Nvidia has, you know, they make custom modified chips for folks and they competed pretty hard. Google did have the TPUs, the Tensor Processing Units, but those have not really performed for the applications in the same way for other companies. This is not my area of expertise.

No, I mean, I'm hearing people talk about this shortage now, and it does seem like it's an opportunity for massive competition. Again, open source and other technologies.

It's just the hard— that hardware is made in like TSMC, Taiwan Semiconductor, and that's a whole geopolitical risk too. And

It's pretty amazing that we've been talking about this invasion of Taiwan now for close to a decade, but that it has gotten more and more heated the last couple years. And that increased tension seems to parallel the development of AI and COVID. I don't know.

Yeah, that is like Taiwan built itself on the back of its semiconductor industry because that was the thing that would get the most protection from the US over time. And that was a good bet. That was a good call on that. But I think even TSMC is saying, hey, maybe we need to diversify a little bit and start opening up fabs in the US.

Oh, and they are. Yeah. India, everywhere.

Yep.

It's absolutely fantastic. So show me your product before we wrap up here and run out of time. Uh, any other Any other like things you're super proud of? I mean, it's super impressive. Um, and I understand now the difference. Uh, yours is doing a lot more inputs in the way it's kind of like Zapier or If This Then That kind of is built into it natively so it can go find more sources dynamically.

Yeah, it can go do that. And then I think, you know, over time, and this is what I was getting to about software in general, you know, You can— you've bought software for your team before, right?

Sure.

And then what happens if the team doesn't use the software to its fullest potential?

We unsubscribe at some point. Hopefully somebody remembers they put it on a credit card, but I cancel our cards every year where I have the cards that you can set the limit on where like I just say, okay, take all the cards down to zero and let's watch our phones ring off the hook as SaaS vendors are like, hey, your card's not working.

That, that definitely works. And one of the, the big, the big shift here is when you move away from software and towards agents, the agents know how to use these tools better than people do. Hmm. And so they're going to extract all the value out of these tools. And that could be APIs, that could be really anything. And you're seeing it across the board in all these little point solutions. But imagine wrapping all those point solutions into a generalized platform. That's That's what we're building.

Makes sense. Yeah, this is, uh, pretty crazy. Well, listen, you got a team of 30 or 40 over there.

35. We're, we're being lean. Yeah.

I mean, uh, you raised some, a decent amount of cash. You got a decent amount of revenue coming in. Are you finding your clients are— this is your, you're too far ahead of what they need, or they're just trying to get, understand this, or now with ChatGPT-3, 5, and 4, they're all like, oh my God, we have to catch up. I know we paid for this. We need to now expand it and we need to understand this because our bosses are breathing down our throats.

It's a mix. So yeah, there's an AI maturity model. Gartner published this model in 2019 and they said, hey, look, we're going to start by companies kind of being interested in different potential with AI. I've taken that and adapted their maturity model. And basically that phase 1 is like the end users are going to start bringing in different AI tools into their day-to-day workflows. So that would be like people using ChatGPT, using our chat, you know, that's also powered by ChatGPT or, you know, using copywriting tools. I saw Canva, Adobe, you know, everybody's—

Notion. Notion.

Everybody's got an AI feature kind of a bulk wizard, which was, you know, I think it took 2 years, and probably the 3.5 model was really the point where those companies would say, okay, we can, we feel okay trusting this and putting this out there for our customers. So that's really phase 1, is, are your— is your team using it? The second piece would be, is your team able to actually scale its usage of AI? And that's not something that ChatGPT is going to really help your team do. Um, but if you wanted to make, you know, repeatable processes, um, build, you know, build actual components of your business. So focusing on the system rather than the manual tasks automation. Yeah. Um, so that's where, you know, we built into that. That's kind of phase 2. And then phase 3 is when you connect, you begin to connect enough of the workflows together so that the data feedback loop starts to really run like a flywheel. And at that point, now more and more of your business is going into autonomous mode. And as you go into autonomous mode, people have to actually build that system out. And it's not the AI team, but it's actually the business users. And so we're seeing—

The process people, the business leaders,

even the functional people. A lot of people went want to actually build these systems out. And so, you are seeing kind of 2 groups emerge. One, they're excited about building more efficient systems, being able to do more projects than their team could have ever done before. And they're the ones that are really AI native, and that's who we build for. And then you have folks that are really hesitant to adopt any of the systems and tooling, and ultimately, Those companies are not going to be competitive kind of in the 12 to 18 month timeframe. They're going to become rapidly uncompetitive. And then pretty much at every company, the CEO is like, hey, AI is here. Like, what's our AI plan? What's our AI strategy? Right. And so then there's a really big gap between the end user tool and the CEO.

Yeah.

Right. And so where we found success with our platform is selling the vision of the platform to the CEO, C-suite, and then the leaders of all those groups. Because they say, okay, well, can you do this use case? Can you do this use case? Can you do— it's like, yes, yes, yes. Because it's all basically just stacking these blocks together. And that'll cover that. I mean, that could be a 10-year timeframe where that really gets adopted, or it could be like 3 years, 5 years. It really depends on the level one of interest, and then 2, it's like, how fast is the underlying technology really going to improve? Right now the bottleneck is people don't even know what's possible with AI yet. So there is some education happening.

I think that's the key, and that's why I just told everybody to start playing with it. It's kind of like mobile. One of the first steps people did in mobile was, or even cloud, was just give everybody a mobile phone. So everybody had flip phones and then like just Get everybody on your team a mobile phone. Get everybody on your team an iPad. If you wanna build iPad apps, they gotta at least have iPad. If you wanna see what 5G and broadband's like, just buy all your executives broadband at home and get them laptops and show them how it works, right? They need to really start soaking it in and playing with it, and then the use cases will come out. And it's like anything else, some people will go too slow, like Microsoft did with mobile, and they'll give mobile up.

They didn't miss this one. They did not miss this one.

They didn't miss this one. I mean, this is like the makeup for missing mobile, right? You think about how badly they— missed mobile and now how well they hit this one. They're like, we're not going to miss this one. We'll pay $10 billion. And then you look at like Mark Zuckerberg with AR/VR, or maybe to a lesser extent, Apple, who has a device coming. Maybe they just also made huge bets and those huge bets are just not going to hit the mark. So it's one of the great things about technology is how dynamic it is. All right, listen, Paul, you're amazing. Really great overview for people who don't know about it and If you're in the business of running a business, I think that's as plainly as you can say CopyAI might be able to help you be more efficient. So go check it out.

We have free plans.

There's a free plan. Go, so go play with it until your free plan, uh, you use up all your credits and then, uh, you go, or what do you do? Multiplayer mode is where you have to pay.

We're doing, uh, right now it's still credits because it, you know, yeah, it's a little pricey for us.

Oh, it is pricey. Yeah. You have to pay. Yeah. I mean, is it, I thought OpenAI dropped the price like 100% or 90% or something.

They did. They, that saved, that saved us pretty good.

Yeah. 'Cause that was, it was going to be quite expensive and they'll do that again, I guess. I bet. Right.

Yeah. Over time it's going to go to zero. And so then you'll have the scale, the volumes will really start to work through the system. And that's another reason that I wouldn't want to focus on the end-user point solutions because you won't have enough volume over time to really propel a business.

Yeah. It's incredible how fast—

Something to think about, you know, something to think about because there's so, it's so easy to create companies and you're seeing this AI hype race, right? But then the durability of a lot of the point solutions is really going to be questionable, I think.

Well, I also think, You have to also want and have the motivation. And when human motivation— this is a very hard reality for most people, especially woke people or people who want to believe the world is fair in some way. Like, it really does break their brain when you're like, okay, now solve for motivation because I can take any course from MIT or Harvard or Stanford online for free right now. And so the world's unfair? And those Ivy League schools have now given all their information out for free. Or the VX100, you can buy a used one and make a movie. And so Hollywood's holding you back, or you could get 3 or 4 of your friends together and make a movie this weekend or a short film for free. What's holding you back? And that's what I learned over time, is that human creativity is all there, but motivation, that fire in your belly, the desire to go do something, sometimes the unhealthy obsession and desire to go do something That's one that we still, uh, as much abundance as we have, can't seem to, uh, solve for. We can't make people motivated to go change the world. Some people are and some people aren't.

My, my oldest daughter, I asked her like, what do you want to do? You know, when you grow up and she's like, well, I want to, I want to breed dogs and I want to do art. I'm like,

sweet.

I said, you could do that now. Like, you don't need to go to college.

Yeah. She's gonna be a dog breeder and make art, and that sounds like a pretty dope life.

She said, she said, I'll breed dogs and then I'll sell the dogs and I'll use the money to buy more art supplies.

And then I'll do more art. And then I'll do some more art. Yeah. And buy more dogs.

Right. So I think if, if I could have one wish, it would be that when we do go update the curriculums of schools, that the only thing that we instill is like that creativity and getting things done that you care about. And if you drive that passion forward, Everyone will be successful because you'll have unlimited tools to make things happen.

There's literally an— a school of, um, academia. So there's Montessori and then there's something called Reggio. And Reggio, one of the core concepts is, uh, Reggio Emilia is figure out what the child is motivated by and then have the curriculum taught through that. In your case, art supplies and art and puppies. And dog breeding. So you could teach biology obviously through dog breeding, and you can teach math and you can teach science through the science of oil painting or sculpture and, you know, the physics of it. You can teach anything through a lens in which somebody's, you know, if somebody's into skiing, you could teach them physics of skiing and they would learn all the physics concepts, but they would be super motivated. And then teaching 'em physics abstractly, they're not motivated. Turns out every kid could learn physics or pick a topic if you just put it in the right wrapper. But yeah, the education system is broken. I just, we should just ask CopyAI right now how to fix the education system. I'm sure it has a great answer.

Well, you can do custom, you know, custom learning plans for people. People are already building those systems out and even like linking to the right YouTube video to watch.

That's awesome. Yeah. If it nails that, it's incredible. Yeah. At some point it'll be like, hey, where— what is this person stuck on?

You know, nobody's gonna be stuck.

You'll do an assessment of somebody, like they're, you know, like some kid's lost in life. They're like, oh, this kid's lost. I'd be like, okay, let's figure out some assessment tool. And then it's like, okay, here's what would motivate this kid. Uh, watch these 5 videos. And it's like, watch this gladiator video, watch this like skiing video, watch this puppy video, and then Listen to this song, and it's like it's gonna unlock their brain and unstuck them.

That'd be awesome.

That'd be pretty great. All right, listen, we'll see everybody next time on This Week in Startups. Bye bye, everybody.