The New Muse: Hobbyists and AI
28:45 Share

You have to keep users in flow. And anyone who's coded or anyone who's gotten into drawing or knows what it feels like to be in flow,

Can an AI help you stay in slow? And then the last bit is that you don't quite know everything that you want. You're still developing your taste as well. And so the kind of building a community for hobbyists versus for professionals, you are still developing your own taste and you want to do with like-minded peers. Yeah, you got to it on the last part. You want to do it with the people who are like-minded. I mean, the internet, that's the thing that's been the best on the internet for two decades now, right?

Welcome to Hallway Chat. It is Fraser. I'm Nabil. We are going to talk today about what was literally a hallway conversation throughout this past week. Let me set things up a little bit and bear with me because it might take a couple of moments. We had a company come in, we won't mention the name, and they were doing really interesting things with AI. A wonderful pitch and then a very rewarding conversation with us as the team afterward.

It was using AI to basically enable a creative endeavor. And it was stitching together a bunch of different models and a bunch of different modalities to make things dramatically easier for people to be able to create things.

And output. And let's just leave it at that. Think of it, tools for thought, creator economy, like generally in the ilk of make things and have a co-pilot to help you and agents to help you make things. That's right. That's right. And it was taking the very common thing that's been discussed over the past year. It was taking things that were inaccessible to many and making it a lot easier for those who wanted to get through it.

Yeah. And you know what? Our conclusion was it's a good product, a great use of this technology. And then we got really hung up on how to think about the product size. And I think that this is the part that was most interesting to me and then led to the conversation. And you had a whole bunch of thoughts on it. And it was your question of like, well, what leads to the larger market? How do you grow the market beyond prosumer edge expansion? Yeah.

And where did you end up with that? Yeah, I mean, what I love about this discussion is that it's exactly what's ongoing. We're not at the conclusion yet. We're using this as a way to kind of navigate and make sense of the world around us. And I love it when we have one of those moments to have that discussion. To back up, the kind of general way that people think about these markets is separating it between professionals and prosumers. You have more individual professional people who can pay and the kind of like large enterprises.

This is basically bottoms up versus top down sales motion. And in that particular market, it just all felt too small. It all felt too much of a tiny wedge that's going to make everybody 8% efficient and just not that interesting. And ultimately helping some large companies' profit margins go up by 1% when they're not even that big of a company. You're not changing the world. You're not really doing anything differently. Right.

So my example was mid-journey as a lens with which to look at this problem. And as we started to unpack this, it just felt like there were a whole lot more mid-journeys around us that were under-explored in AI. And that's, I think, when we started to really pull away things at it. But wait a second, like...

Midjourney is a prosumer tool for illustrators, isn't it? Like it's allowing people on Fiverr to create imagery for their slide deck, for their workplace. Isn't that what Midjourney is? Isn't that what these text-to-end models are? First of all, I'd argue that if you want a cheesy slide, clip art slide, you should go to Dolly, not Midjourney. Come on. Whatever.

Sorry, sorry. How dare you? How dare you? I think it is... There are competitors that are focused on those markets. There's lots of competitors in the industry fusion model market. MidJourney is...

focused on net new creators. I think we should just remember that there's some very unique things about MidJerdy. People mostly talk about how it got to scale at Discord, which we can talk about another time. I think the most important thing is that David thought their original customer was going to be like

14 year old meme maker artists that were going to grow up and be the next wonderful AI artists of the world. And that's not what it's turned out to be. It's 30 and 40 year olds that are semi-professional non-professional who are doing this because they love it. They have a new comic book that they're going to get to build for the first time. They're exploring new architectural designs. They're trying to put up their own pieces of work for their own love. Like it's,

Less close to Activision and EA making concept art for a blockbuster video game and much closer to scrapbooking, which is like a three, four billion dollar market or gardening or knitting or amateur indie video game builders. It is people who love the craft and are exploring it for its own love. It's the hobbyist market. It's hobbyists, right? Yeah. Yeah.

I think we often listen to the loudest customer in the room when we are trying to do customer development in 2024.

as a good product builder, as a good founder. The loudest customer in the room is almost always on the professional side, the large enterprise side. And then we kind of started to learn about the prosumers, especially in the B2B SaaS, bottoms up Canva style world, eight, nine years ago, move everything to cloud, you get to reach people with Sigma or Canva that are inside of the organization, but you're selling bottoms up, not top down.

maybe the solution is like, if AI is really going to enable a thousand X more people to create, then where you start from is the people who want to create that aren't creating today. And the best way to think about those people is the hobbyist market. Yep. On the, on the text to M side, it's, it's interesting. Maybe the,

A year ago, there was a lot of, I don't know, dinner table conversation is Adobe at risk. There will be entirely new professional tools. And it may just turn out that the professional tool for text to M and these image models is actually Adobe. And it's just a sustaining technology that gets added into the super complex environment.

And it may be that the prosumer market here is actually Canva, right? It's like, hey, if you want to make a slide for your presentation or what have you, you go to Canva. And I don't think many people were talking about Canva.

this behavior last year that there's an entirely new market, which is massive. Like, I don't know. I was stunned when you shared how many mid-journey images you've created on your account. How many? Yeah, I'm well over, I don't know what the number is, but 30 or 40,000 mid-journey images. And that's not even, I love it. That's not even counting the Leonardo AI or ideogram or all the other images that I've been working on just trying to train those products. Like, yeah, I have side projects.

and I have started to learn that particular language. And I certainly wasn't going to do a year of mid-journey coursework to become a mid-journey artist the way I might have had to learn how to do figure drawing or learn coding or anything like that. It's a situation where you get control over the model over time.

iterating on it and working over time. But the floor is quite easy. It's a easy to learn, hard to master kind of arc that you really want when you're trying to build a new skill. And it's just important to recognize two things. One, that to give David credit, he recognized that audience early enough and adjusted his vision to try and meet that audience where they were.

And secondly, or overarchingly, he internalized how different that product would be for a hobbyist type of customer. And you can only think about if I'm a very casual user and I just want clip art in two seconds and I have no expertise at all, then Midrini is a terrible product for me. I should be using, frankly, something like Golly.

And if I'm a prosumer user who maybe has used Adobe some and understand that interface, then maybe something like Leonardo AI is the right interface. Sure enough, I think Leonardo was acquired by Canva recently because it nailed it for that. But yeah,

If I've never used Adobe, Leonardo feels way overwhelming as a user interface. And if I don't quite know what my style is yet, then the mid-journey Discord server where I'm seeing everything else going around and we're trading prompts and it feels like a little community there, that becomes a feature. But hobbyist users are not the same as consumers.

If I'm a super casual user just trying to very quickly get something in a PowerPoint, suddenly all this community stuff is a bug. And we could see the same thing happen in many other categories, which is, look, if I'm going to make music,

We have this memetic, very casual music product, which is like I want to type in a prompt. That's kind of served by the Suno and Udio and Refusion and those companies' use cases. That's kind of similar to the DAL use case. I wonder what the mid-journey product for music is. I kind of say the same thing about podcasting or many other things we make. I think everybody should make, and you should make it easier. For me, it's also just like an interesting puzzle to try and unlock because...

We have less of a product playbook and a go-to-market playbook when it comes to the hobbyist use case. And yet that might be an area where AI might have the most effect. I think what you just said there is really insightful and it's worth us spending a couple of minutes on because this is a market where it is what you create rather than how you do it in my mind in terms of where the joy is sparked for people.

If you wanted to be a musician previously, you learned the fundamentals of music, you learned how to read sheet music, you had to learn scales, and then you'd like slowly build up on how to do it to the point where you can then create the music. Similar with like visual art, right? Like you have to learn brush techniques and color theory and everything else. The opportunity is much more emphasis on the conceptualization, right?

rather than like the craft of doing, if that makes sense. So let's assume that if we're trying to, if we're a founder, we're trying to find new other markets where this is going to occur.

is you start from the place that people spend time in their hobbies already. And you basically, you find the places where people go for their own joy. You start from the places that it's uneconomic. Right. And some people will end up doing it forever for uneconomic reasons, which is wonderful. And then some of those amateurs will turn into the next professionals. You know, like I often think we have like a mass professionalization problem, by the way. This is kind of a little bit of a side, but like I was reading...

Robert Morris is the power broker and he talks about the shift in politics. We're not going to talk about politics here too much, but he talks about how like it used to be that like something like half of the house of representatives and three quarters of the Senate or something along those lines didn't even have a college degree.

And that's just completely impossible to happen right now. And the transitioning is a little bit like home-cooked meals turning into restaurant dining, which is like, it can seem more polished, but we're losing some of the authenticity and some of the diversity of perspectives that leads to better innovation over time. So I think it's an incredible fertile field. This can feel like a different founder skill because if I start from joy and I get to things like,

people make their own video games, whether it's uneconomic or not. That used to be a sign of a negative because you're like, oh, it's a red ocean market and there's a ton of competitors and one more video game, what does that do? And instead that becomes an opportunity to enable people if you can find a way. They'll make video games, board games, knitting, gardening, not all of these are convertible to AI just yet, but those are the

the fertile ground, right? The pushback is, well, and this is maybe a proverbial straw man, if the AI is making it, then am I actually getting

good at it? Am I actually building taste in it? Or am I not doing grit work and so I will never become world class because I didn't actually learn to draw our code? I think that's so, that's the proverbial straw man and it is so like inane to even be asking that, right? I don't

I feel like it's a perspective. You know me, and I know you're looking at me like I'm ridiculous for you to bring it up because you know me, but I think it is a worthy subject to spend 60 seconds on because it would be a common conversational worry. Oh, 100%. If somebody's doing all your AI writing for you, you're not going to get an amazing author. I put that in the bucket of the people who are like, actually, Frankenstein's monster was the author or whatever, like that pedantic bullshit is.

No, because you are not learning the how. You don't know anything about brushstrokes. You don't know anything about color theory when you're grinding on your 27,000th mid-journey image. But you are certainly appreciating and having joy in what it is that you're creating. And I assume that you have learned how to speak

about what it is that you like and what you dislike in imagery in a way that you hadn't before. You're probably learning to critique things, even if you're self-taught. I think it's just so different from what you had to do that that conversation and that question that you asked is completely fair. But the answer is just it's no longer about the process traditionally of how you got there.

Like, yeah, there's a difference between art and craft.

and they have been merged together because they involve them, each other, but they are separate things. And we are separating those things now. And learning to make mid-duty images is similar to learning to play an instrument or learning to write. The more you practice it, the better you become. And the more you develop your unique style, the more you build your own sense of your own voice. And all of that stuff still happens. And in fact, I would argue that

The fact that you can do it 100 times or 1,000 times faster should give you those proverbial 10,000 hours of practice even faster, which was not about becoming a good guitar player, although the 10,000 hours thing from Malcolm Gladwell's Einstein is provably bullshit. But the general theme of practicing more helps you build your own sense of taste, and your own sense of taste and the voice you want in the world is actually the goal, right, in whatever way it is. Yeah.

Yeah, for sure. Maybe there's a really simple way to look at two different product categories to like really firm up how we think about it. And there is broadly a discussion around AI writing tools. But if you like go below the surface, there are...

tools that are clear homework helpers and we'll talk about that in a moment and then there are tools that are created for writers and let's put some space between those two where homework helper with air quotes around it you write the essay topic you hit tab tab tab tab tab and it just auto completes paragraph after paragraph including now cited references as if it's like you you've written it

There's little to no thought in that type of a product for the person writing it. And it's clearly not being built for somebody who wants to lower the activation energy of writing, let's say, a manuscript or a novel. It's not a thought partner. It's trying to finish the work for you. That's right. That's right. And then there's stuff like pseudowrite where it's saying, hey, writing is a creative outlet for

And there's now new tools that can help you overcome some of the challenges that are present when you want to write a book, a play or what have you. And this like clearly crafted in that in that direction. I agree. And that is a good example of a thing that people do for joy. Like my father has been writing his memoir for the last couple of years. I don't think he's going to be a bestselling author. That's the goal. He's doing it because he loves the process and he loves the work. Yeah.

A third of all Amazon Kindle sales are to self-publish work at this point. So it is a big business in aggregate, and it's still a large market in aggregate. But it's very different than the influencer process, right? The kind of like creator economy influencer market is I'm going to go do this thing, and then I'm going to stream on Twitch. But I'm mostly doing that to see if I can be...

a big star later on. Like, I'm going to Hollywood to see if I can become a great actor. And that's remarkably different from these markets that are hobbyist markets where the output is largely for its own reward. Yes, you want feedback from peers. Everybody loves good respect and love.

and accolades, but there isn't, except for in the most fleeting, tiny way, an expectation of stardom or expectation of trying to be the 1%. You were doing it for the joy of doing it. Right. And in that case, if you've written your autobiography or what have you with one of these tools, you might not have appreciated

appreciated the how as much as if you like grinded through all of the books on how to write well and everything else along those lines. I think there are challenges and opportunities to the products when you're trying to build for hobbyists. And one of the reasons why it's important, of course, to focus on a specific customer is you just build different things for different customers. They're just loud about different things. Right. And when we talk about making, writing is one of these categories, but

But music, for sure, with companies like Suno and Udio and Refusion and others, music is a thing that people do just for the love of it. Design, there's Canva and Fig. Podcasting, there's Descript and Captions. And all of these over time are going to separate out, ideally, into finding niches of their own customer and then do very different things. I think for hobbyists...

The benefit is that to building and doing is you have a permission to fail with relatively low stakes when you're building in it. That gives you a capacity for risk and innovation, which is the thing I love about it, that you maybe wouldn't have had when you're like, life depends on it tomorrow. And that presents in a way a hint to what these people bump into, which is that nobody's making you do this tomorrow. Right. So any friction, any problem,

can derail you completely. And that doesn't mean you have to gamify it. It absolutely doesn't mean you need badges. But it does mean that you start from an activation energy situation. You have a blank page problem. Instead of having to stare at the blank page and think about where I'm going to start, can I just sketch out the characters that I might think about or how I want an ending to happen if I'm writing and work my way back? I think the second bit is you have to keep users in flow. Every time you fall out of flow...

from anyone who's coded or anyone who's gotten into drawing or knows what it feels like to be in flow.

can an AI help you stay in flow? Like a great example, this writing music can be torturous. I think there is a product out there in AI music for somebody that helps you keep you in the music, writing music, authoring flow the whole time. And then the last bit is like, this is what I think discord was the unlock for mid jury, which is that you don't quite know everything that you want. You're still developing your taste as well. And so the kind of building a community for hobbyists is,

versus for professionals. You want to see tons and tons of real-time examples because you're still very open to input. You're not a pro that's been doing it for 30 years that knows your thing is just trying to get published. Just show me marketing tools. Get me a publisher or whatever it is. Go find me a studio gallery to put my thing in. You are still developing your own taste and you want to do with like-minded peers. You got to it on the last part. I think there's another...

piece of it. And that is like-minded peers. Like, sure. Yes. It like teaches you, you can explore, you can refine, you can learn, but it's with people who are hobbyists who share your passion. I want to go to the knitting club down the street. I want to go to the garden tours and appreciate the joy of other people who share this passion. That's like built into the product. That's so insightful. I used to think that music

I thought that there was going to be AI-generated music that would reach the top of the charts. I still think that, right? I still think that. I looked at, we met with a whole bunch of different music companies, and oftentimes the pitch was basically that. And you're like, I don't know, is that a, I don't know, is prosumers? Sure, I'm not really sure what's going on here.

And we saw one where somebody pitched it as like a music social network. And you saw, like, I was struggling with it in the moment. Now I get it. I get it. If you're lowering the activation energy and helping people, you know, find joy in this as an outlet, you do. You want to do it with the people who are like-minded. I mean, the internet, that's the thing that's been the best on the internet for two decades now, right? Yeah.

Yeah. It also fits into this broader picture, which frankly, I see it across all these economies in any market that develops over time, which is eventually the middle drops out of most markets. Like you think about Hollywood movies and like the $100 million rom-com has just completely gone away. You get, you know, a billion dollar spent Avengers sequel, the 56th sequel, or you get long tail YouTube creators and that middle goes away.

I think it's happening in venture capital as well. Quite frankly, I think it's going to happen across most markets. I think you're certainly seeing it in music like that. Kind of like, can I be a career musician who makes a million dollars a year, but not 50 million? I'm not trying to be Taylor Swift. Right. And I'm not just trying to be a YouTuber. Like that market is very, very tough when markets get efficient. We can lament that middle of the market going away, but it is just a fact. It's a fact of the way all of these markets trend. And that's going to happen

with or without AI, you can look at, again, you can look at most creative endeavors and in video games, in music, in,

movies like they are all trending towards that middle of the market dissolving and you're either in long tail or in the peak. I think there's a lot of effort and time that's been spent on trying to make the long tail chase stardom when and think about instead how they should be doing it for their own love of the work and whether they can

make enough money to keep doing it for the love of the work and good things will come out of that. I think you build very different products. You don't worry as much about marketing distribution and making you a star and you don't create feedback loops to train that you want stardom. And I think a major competitor

not your job. The major competitor is what else you might do in the evening or on weekends. Like this is fighting with Netflix and Fortnite. That's your major competitor. You want someone to sit down and make art in mid journey. So they're at least making something instead of watching yet another Netflix show. That's the goal. We like, if we can make more people who are making,

making things out in the world and pushing out their voice, starting another podcast, writing another book. It doesn't mean it has to be for you. It doesn't mean it has to make you a billionaire, but it will help us like move the whole conversation of culture and society forward. Yeah, super interesting because it's not that you're

Mid-journey use is coming at the expense of Canva. It's like, or like you're not, you're not saying, hey, I'm not going to go to like that $5 marketplace to get my needs served here. Or even my Spark job. Like it doesn't eat into my time with founders or my time talking to you, right? Yeah, your aspirations is not to become a very well-regarded illustrator or artist.

you're doing it purely for the fun and the joy and the escape that it offers you. That's why I love it for my kids too. I am thinking about their time and I would much rather they can make a board game or a video game or make art in mid-journey or do any of these activities than, I don't know, play Fortnite for longer. Right. For sure. I get it. How interesting. All of a sudden, these technologies are allowing creative endeavors through the computer

With an audience that never would have been able to turn to it, like any of it. That is so inspiring. Yeah, it is inspiring. It is the opportunity for many more people to have their voice heard in a way that reminds me a little bit of the kind of like, no, you know, you know. No, because I mean, sure. I'm sure you share some of your mid journey imagery with some people.

But you don't do it to be heard or seen or discovered. That's fair. That's fair feedback. It is okay to do it for its implicit joy. The people who turn wood and do carving or build furniture for themselves or paint, the people who are completely exposed to these creative outlets are doing it for themselves for the most part. And now AI is allowing creative endeavors...

en masse through software. That's right. Yeah. You don't have to worry about the how anymore. It's an important part and a macro sense of like how I imagine how we innovate, but it's also like an important part how we live. Let's not skip over the fact that feeling like you are making something and feeling like you are executing on something and putting something out there in the world that you feel pride in the making of when you look at it, it's just a part of living a good life.

100%. So you're not going to get this analogy, but it makes it like bees wiggle to transcend to other bees, like paths and so on and so forth. That's their language. And then 30% of the time, apparently, according to science, like bees wiggle for no evolutionary reason at all. Yeah, you're just wiggling. Yeah.

You're just wiggling. I did not anticipate me getting here, but I feel so optimistic about all of this now in the sense that before to reach creative expression really required mastery of the how.

And I know that I keep coming back to this. Yes, that art and craft was a merged thing. That's right. That's right. But it was hidden. Like, it was hidden behind the how. And I would love to create music. I've never invested into the how in a way that is, you know, rewarding enough to then have the output be something that I feel good with. Yeah.

It would never be like, I don't have any aspirations to have other people enjoy it. Like, I just want to be creative. And there's now opportunity across so many different mediums for that. By the way, this is an interesting, another orthogonal dimension to talk about this, which we can maybe end on, which is like, and this is not to say that craft doesn't have its own sense of merit too. No, not at all. Right? And in fact, there's a whole other AI product set

which is about being your tutor, which is about making you good at the craft. Go Try Simply Draw, which just launched. There's a whole set of like how to play guitar or how to become a good writer or how to solve a math equation that is about the craft of the work. But that's not what we're talking about here. We're talking about a hobbyist who...

is already has a vision of the thing they want to make and they just haven't learned how to craft it yet. And can you help them along the way? Oh, that feels so good. Actually, what you said earlier feels so great. It's that historically that the creative outlet is meshed

With the craft and the how to do it. And certainly some people love that component of it, but not everybody. And now they've been separated and you can choose. You can choose to be creative without having to master the traditional how.

And that feels so good. Yeah, I agree. It will not be universal. Not every category of AI products is going to have this. It's almost three categories now that we're talking about, which is first, where's your customer sit? Are they prosumer? Are they consumer? Are they hobbyist? And it's just really talking about that maybe the larger opportunity in some categories, although not every category is hobbyist. And then second, in those categories, particularly for hobbyists,

The art of it, the expressiveness of it is probably the most important thing versus the craft. And then lastly, like there's probably a whole nother conversation that we can do on people that are helping with the craft. Yeah, for sure. For sure. So great. Well, thank you for this. Yeah, no problem. Let's be done for today. Talk to you later. See you.