AI Rights and the Future of Publishing with Created by Humans Founder Trip Adler

Ricci (00:01.69)

Hi and welcome everybody to the Written Word Media Pod. Today we have a great episode for you. We are joined by Trip Adler, the co-founder of Created by Humans and also the co-founder of Scribd. Trip is here and he’s gonna tell us all about his new endeavor. Welcome Trip.

Trip (00:22.882)

Thank you, it’s great to be here today.

Ricci (00:24.657)

Great to have you. So let’s just start at the beginning for those people who don’t know you, who don’t know Scribd, maybe you can just give us a quick background of where you’ve been and how you’ve worked in both publishing and technology over the last few years.

Trip (00:43.99)

Yeah, sure. So before Created by Humans, my last company was called Scrib. This is the first company I started. And there we pioneered the subscription model for books. We worked with publishers and authors to create this new model for the industry. And we built up a pretty big business, over 200 million revenue, over a billion annual visitors. And there we worked to kind of…

bring this new model forward and connect authors and readers in new ways. I did that for quite some time. Then when this AI thing came along, had a lot of folks in the book industry reaching out to me, asking me questions about what to do about AI. Meanwhile, I had folks in Silicon Valley asking me questions about the book industry. Things started in a tense situation between the AI industry and the book industry with lawsuits and everything.

people were coming to me asking me for advice and my thoughts on it. it just felt like this is a problem I really want to try solving. From Scriabank, I got know a lot of people in the book industry. I got to know people in the Silicon Valley. And I just wanted to try to bring these two worlds together. So at Created by Humans, we’re trying to build a model that connects authors and connects the industry. And that’s what we’re working on now.

Ricci (02:06.456)

Okay, great. So how long ago did you start Create About Humans?

Trip (02:12.01)

It’s been a little more than a year. it’s a brand new company. We’ve raised venture funding and we’ve got a small team, we’re working away. We’re making very good progress. When we started this, we really had very little idea how to solve this problem. We just knew that there was a lot of conflict, a lot of confusion, and we figured someone had to figure out how books were going to fit into an AI world. So we just started working away at the problem and we’re making very good progress.

Ricci (02:13.904)

Okay, so really new.

Ricci (02:42.273)

great. So do you have a sense of like what the mission is or what you think created by humans will become or what some of the goals are?

Trip (02:51.692)

Well, the mission is to protect human creativity and make human creativity thrive in an AI world. A lot of creators are kind of worried that AI is going to just replace human creativity. And we think that is not going to happen. We think that’s a ridiculous idea. But we do think, though, a model needs to be put in place that protects human creativity and creates a business model around it. So that’s what we’re building.

And we’re doing this by creating a marketplace where we call AI rights. So AI rights are a new type of right that content holder, rights holders can protect and license out in this AI world. We’re doing this for books to start. We do have ambitions to do this for all kinds of content. We think this will work really well for music and for video and for all kinds of human creations, but we’re laser focused on books at the moment.

And we’ve created this concept of AI rights and we’re creating a marketplace for them.

Ricci (03:50.222)

Great. And so the concept of rights, that’s not really a new concept in publishing. So can you talk a little bit about what kind of rights already exist and how AI rights might fit into that existing ecosystem?

Trip (04:04.246)

Yeah. I mean, rights are they they’ve been a concept in publishing for a while and it feels like people in Silicon Valley don’t think about that too much, right? It’s like, so that’s why we need kind of a new framework for how to how to bridge the world of rights in the world of of AI. So, so, so, rights, we’ve we’ve we started thinking about like, you know, what are our rights? And we we realize there’s actually three distinct types of AI rights. And we’re building our business around these three types. So the first is training rights. And this is the right to use a piece of content for AI model training.

you know, to, to, date, you know, like exactly, I’ve argued it’s fair use, and it’s still kind of playing out, in, in court. but, but we think that there, there is, there definitely are some companies that want to license training rights. So, we’re, getting some deals going there and we think, you know, gradually we can get it, get a, you know, a real market going for, for AI training rights. the second category of, of AI rights is what we call reference rights. And this is the ability to, make books available to AI systems via RAG. So, so RAG is a new,

a term that’s used quite a bit in Silicon Valley. You can think of it kind of like search. So sometimes when you ask a question to chat GPT, it pulls the answer from the train model and sometimes it actually pulls it directly from the source. So there’s been lot of deals for RAG, a lot of news deals, but we’re trying to get more of these deals going for books. And we just signed a really…

Ricci (05:26.235)

What does, sorry to interrupt, what does RAG stand for?

Trip (05:30.274)

It stands for retrieval augmented generation. So for example, if you ask a question of chat GPT and the answer is on page 200 of a book, we could actually kind of get that answer directly from that book and serve it up to the user. And that’s just different from training. So these AI companies are looking for data for RAG and we’re basically building the framework to allow it to be possible for books.

Ricci (05:56.726)

Okay, so training, rights, red, rights, and then what’s the third one?

Trip (06:00.6)

Yeah, we call them reference rights, a little over a bookish word, but yeah, rag rights, same thing. Yep. And then the third one is we call transformative rights. So this is the right to take a piece of content and transform it into something new using AI. So, you know, we’re working on a deal right now where a startup is going to turn books into virtual worlds. Or this could be turning a book into a chat bot. You know, this could be just, you know, turning books into new things using AI.

Ricci (06:02.965)

Okay, rights, that’s…

Trip (06:29.57)

And a lot of this is already kind of happening in sort of an unlicensed way, but we’re building the legal frameworks that startups can now, startups and big companies and all kinds of companies can actually get the content they need and get it in a legal way and have confidence in doing this in a legally correct way.

Ricci (06:46.059)

Got it. So if for authors or publishers who are out there, what would the benefit be to now allowing the AI companies to access these rights or the ability to sell these rights to the AI companies?

Trip (07:03.896)

Well, think one thing is it’s just a new way to get your books out there and to reach readers in new ways. mean, consumers’ expectations are changing quite a bit. They’re all expecting things to be AI experiences or something things to be interactive and dynamic and personal. So this is just a new way to get your books in front of readers in new way. And it’s also a new business model for your books. If you look at the history of books, was…

There was print books and there was eBooks and there was audio books. We think that AI licensing is the next big model for the industry. So we think it’s going to be an important new model. And the fact is that these AI, a lot of AI products are already using books in various ways and it’s already happening. So we’re just trying to create a legal way to do this where authors can maintain control and get paid for it.

Ricci (07:56.109)

So I’m glad you brought up ebook rights because that’s a great example of a type of right that 20 years ago maybe a little longer didn’t really exist, right? So you had whatever rights you had and then ebooks came along and then you now, a lot of the publishing contracts didn’t have anything referencing ebook rights so when authors published into ebook they were now able to sell those rights. I think some of the other types of rights that are out there are also language rights or translation rights so you can sell the rights to your book in other languages.

Trip (08:04.814)

Mm-hmm.

Ricci (08:26.201)

And now we would have AI rights. So this is another way, another right that you can sell and another way to monetize your box, really diversify your revenue stream. Am I hearing that right?

Trip (08:37.142)

Yeah, exactly. it’s taken some time just to kind of define what these are. think there’s been like lot of, know, AI came along so quickly that there’s been a lot of confusion around like, you know, how books are, how books are they going to connect to the world of AI? So we’ve had to define these rights. We’ve made very good progress on that. So, and if you’re an author and you’d like to monetize and take control of these rights, we’ve built the platform to make that possible. So we encourage you to come, come sign up and try it out.

Ricci (09:01.804)

So let’s talk about how that actually works. if you’re, let’s just start with actually being an independent author. So let’s just say you’re an author and you publish directly through Amazon and the different retailers and you have a hundred percent ownership of your rights and you were interested in exploring created by humans. We’d go to your website, then sign up for an account and then what happens?

Trip (09:25.612)

Yes. So we’ve made it very simple and self-serve for authors. So you come and you sign up, you create an account, you verify your identity to make sure it really is you. You can then claim your books. You can select the AI rights you want to license. And then we have a licensing agreement, which you can sign in the product and that license your books to Create By Human so that we can then go license them out to AI companies. If you’re still trying to figure this out you’re unsure, you don’t have to sign the agreement. You’re welcome to sign up and just play around with the product.

You can choose to sign the agreement later. And you don’t commit to any deals until the deals are actually presented to you. So you can actually approve the terms and make sure you’re comfortable with the company and the terms. Our whole idea is to give authors full control so they can decide exactly if and when how their books are used by AI companies.

Ricci (10:13.067)

Interesting, okay, so you would sign the agreement with Created by Humans, but you then get to opt in or opt out of every deal that you would then bring to the authors. Okay, great, and then what about exclusivity? Like if an author comes and signs something with you guys even before the deals are being presented, and then they hear about some other deals somewhere else or they approach directly by somebody who wants their AI rights, how would that work?

Trip (10:24.875)

Exactly. Yep.

Trip (10:43.162)

It’s all not exclusive. We’re just trying to get this model going and get a business model going for AI and authors here. We’re supportive of all kinds of deals. You can sign up for Created by Humans and still license other places. It’s not a problem at all.

Ricci (11:03.283)

Okay, great. So it’s a non-exclusive and authors get to opt in or opt out. So can you give us some examples of the types of deals that you’re bringing to authors or the types of companies that you’re working with?

Trip (11:16.494)

Yeah, well, they all fit into these three categories, rights. There’s kind of three distinct types of deals. There’s trading deals, there’s reference deals, and there’s transformative deals. And the types of companies are, they’re really, they’re all over the place. There’s interest from the big AI companies, and we just signed a really exciting partnership with one of the big names. There’s also…

There’s also a lot of startups, a lot of startups want to do things with books and they’re looking for the rights to do things with books. There’s a lot of startups, like a lot of different startups. And then there’s also kind of mid-size AI companies. There’s even been governments. We have one government agency that’s very interested in training their own models, looking for books for that. So, yeah, we think there’s just all kinds of companies and it just, there’s gonna be more and more over time.

So it’s really exciting just how many different opportunities are out there that you wouldn’t even think of. And that’s why it’s such an exciting opportunity for books is there’s just so many more, there’s so many new ways to monetize books just weren’t possible a few years ago.

Ricci (12:26.292)

So let me ask the cynical question, which is, up to this point, a lot of the big AI companies have not signed any kind of rights agreement and they basically just use the works. Why do you feel like as we head into the future, these governments and smaller companies and the AI companies are going to be willing to pay for these rights?

Trip (12:49.176)

Well, there are companies that want to pay for content for training. I there’s certainly quite a few deals happening. So there definitely is market there. And I think it’s just a matter time before these big AI companies also want to pay for training rights. It’s just been a little bit legally complicated because there’s these big lawsuits going on. So the companies involved in lawsuits are reluctant to license training rights. But I think as these lawsuits play out, that’ll start to change and they will license training rights.

And that’s also just trading, right? mean, these companies are still looking for these reference rights and transformative rights too. So I think it’s just, it’s gonna take some time to play out for things to sort itself out, but there’s deals happening now and I think it’ll just get better with time.

Ricci (13:32.681)

Yeah, I do think the legal backdrop is very interesting because one of the reasons why, as we move into the future, these companies might buy rights is it’s a way to de-risk, right? Like they don’t want to have to be open to continual lawsuits and some unknown settlement down the road, whereas if they buy the rights upfront and they do everything legally, and this just becomes part of their business model as a corporation or a startup or government, it really takes the risk away from using.

the data in the box in order to get either the training models or the training rights.

Trip (14:09.858)

Yeah, I mean, if you look at what just happened with Anthropic, I mean, there’s a very big payment to authors and we think that’s just the beginning of a lot more to come. And once these lawsuits play out, then I think these companies are gonna wanna do this the right way. And I mean, to be fair, this situation, there was just no structure around AI licensing when these AI companies started. And now they’re becoming big businesses. It just needs to be figured out. And fortunately, we’re just making really good progress in creating this framework and creating a product to make it possible.

and we think it just, you know, matter of time before the sorts of stuff out and becomes a really great new model for, for authors.

Ricci (14:43.942)

Yeah, the anthropic lawsuit and settlement has been obviously a big topic of conversation within the author community. And for those, do you want to just give us like the two cents of how that lawsuit played out and what the settlement is for authors?

Trip (15:01.55)

Yeah, well, the settlement is for a total of 1.5 billion, about $3,000 a book. It turns out that that settlement was really just about the piracy because the books were pirated in the first place. So it’s effectively just a payment for the piracy that happened. I think this question of whether training is fair use is still very much an open question. I think that that’s still going to play out. There’s still many more lawsuits coming and I think those are going to gradually determine whether training is fair use.

I think it’s looking good that it, I think this, it’s probably binary whether it’s fair use or not. It’s going to be some sort of kind of middle ground state, but I think it’s looking good for authors that there still will be a paid model for this over time. It’s going to take some time for the lawsuits to play out.

Ricci (15:48.711)

Yeah, and so $3,000 per title and then if you have a publisher that gets split between the publisher and the author So let’s talk about authors who may have publishers How how can this work for them or can it?

Trip (16:03.66)

Yeah, so we support both indie authors and authors and publishers. When we started this, we actually went to publishers first and said, we’d like to try to solve this. then publishers were the first ones to say, well, these rights haven’t been defined. If anything, they’re owned by the authors. And then they directed us to the authors guild and we formed a partnership with them. worked closely with them. But we support both situations. If there’s a publisher, we work with both the publisher and the author. We don’t license anything until

till both publisher and author are signed off. And that’s a lot of what our platform does, right? Is just sorting out when there’s multiple rights holders, we can sort those rights out. And we do the same thing if there’s multiple co-authors, we also need to get sign off from multiple co-authors. So our platform is designed to allow kind all rights holders to sign off. And then we only license once we have approval from all rights holders.

Ricci (16:56.505)

So if they’re publishers who are listening to this, they can come to your platform as well. They can sign up as a publisher and put in their catalog of titles. And then their authors would come in as well. And both parties would sign off on any deals or agreements.

Trip (17:13.6)

Exactly. Yeah, we do partnerships with both authors and publishers. It’s currently only self-serve for authors. With publishers, we would need to kind of sign a manual deal, but that’s fine. We have a head of publisher relations and we’re happy to form a partnership.

Ricci (17:31.205)

Okay, great. So let’s go back to the money. How much money do you think authors could expect? I know this is gonna be a tricky question, but how much money could authors expect to make from licensing their rights? And let’s just focus on the short term, because we don’t know what’s gonna happen in the medium.

Trip (17:51.606)

It’s hard to say. The deals that we’ve been working on are really very different. One of these deals, it’s a really high price per title, which is really nice. We’ve also had offers at very, very low prices per title. We have quite a few deals where there’s a variable rev share component, where your payouts would go up over time as the licensing business grows.

So there’s really just a large variety of deals and it’s really hard to say. I think it’s just going to be as we do more and more of these deals, market prices will start to gradually form. And then I think in the long term, it’ll just become kind of a new model. I do think that authors should probably set their expectations properly that

that it’s likely gonna be a lot of companies licensing their books, paying a relatively smaller amount. I think that’s how this becomes a big model. More than having just one or two companies pay a lot, it’s more like getting, in the long term, getting 100 companies to pay a smaller amount for their books. So think that’s more likely how this plays out over time.

Ricci (19:05.574)

Can you give any kind of range you said there was like really low and then really high? Any kind of ballpark ranges at like 50 to 5,000? It is a 50 cents to $50 like just to get a sense.

Trip (19:18.374)

I mean, at the high end, we’ve seen tens of thousands per book. So it could be quite high. At the low end, mean, we’ve had, we’ve seen some low ball offers, but those are not deals we’re taking to our authors.

Ricci (19:36.004)

And then on the higher priced offers, is there a particular type of right that tends to be more lucrative?

Trip (19:46.062)

TBD, I mean, we’ve seen some interest in paying high prices for training. seems like training is kind of split. Some people don’t want to pay anything and some want to really high prices. I think for transformative rights, I think that’s also going to be pretty high prices. Yeah, I think it just, it depends. It’s kind of, it’s playing out. I think it’s really kind of early to say.

Ricci (20:10.095)

Yeah, okay. So it is, it’s early and everyone’s learning as we go along. One of the concerns I hear from authors or one of the anxiety points is that if they are, say they sign a rights deal for transformative rights or training rights, and then when they see how their book was used, they’re unhappy and they wanna hit the undo button or they feel like they’ve lost control over their voice or their work.

How do you address that concern?

Trip (20:43.01)

Well, this is, I mean, this is what we’re trying to solve is, you know, we want to create a model where authors are in full control and they can decide if and how their books are used. I mean, all the deals we do, typically the author can opt out over time if they’re unhappy. There’s, you we put restrictions in place for what can and can’t be done with the books. So, I mean, these are the kinds of things we’re working out.

And we think as we kind of do more of these deals, we’ll start to kind of create standards on sort of what can be done and not be done with books.

Ricci (21:23.429)

Yeah, knowing the anthropic settlement, believe part of the settlement is also that that anthropic has to remove all of those books from its training set. Am I correct about that?

Trip (21:34.19)

That’s my understanding. They have to delete the files.

Ricci (21:37.762)

Right, so there is precedent here where if an author wants to pull out or if you do want to kind of undo or remove your book from training, at least in the anthropic model, they’re saying that’s a possibility. Because I think a lot of authors are like, once it’s out there and the training models have it, isn’t the cat’s out of the bag. There’s nothing that I can really do to put it back in.

Trip (22:03.532)

Yeah, I mean, there’s a lot of opinions on this. There isn’t really a clear consensus. But one thing is that these models do need, they need to get new data to train each new model. And they’re typically releasing new models over time. So even if a book was used to train an old model, it’s just a matter time before those models are outdated.

Okay, maybe delete this answer. He stumbled. What was I gonna say?

Ricci (22:31.929)

Yeah. Yeah. Let’s do it again. You can go again.

Ricci (22:40.046)

that models need, they constantly need to be fed recent data, Or more up-to-date data.

Trip (22:47.318)

Yes. So the fact is that these models, each time they train, they need data for each new model. So even if one model was trained on a particular set of books, that model will just over time become outdated as it gets upgraded to newer models. And these models need new books each time they train. But the fact is that these models have already trained on most of the books in the world. So it’s very likely that your books have already been trained on by many AI companies. So at this point,

The train has left the station. as an author, kind of need to sort of get on board or else, or else you’re kind of just gonna, your books are going to be trained on anyway.

Ricci (23:27.013)

Okay, so last question on just how this all works. So the deals are brought to the authors if they opt in, they know exactly what the terms are going to be, how much they’re going to be paid for their title, what the term of the licensing agreement is. What about created by humans? Do you guys make money? Are you for profit? you nonprofit?

Trip (23:51.982)

We’re for profit. We take a cut of licensing fees. So we take a 20 % fee. And yeah, we’ve been able to raise venture funding as a for-profit business. But that said, we’re also a very mission-oriented company. And we’re trying to solve this problem in the world. And we hope that being as a for-profit business, that helps us solve the problem.

Ricci (24:14.808)

But you only get paid if the authors get paid is what I’m hearing. Okay, great. I’m trying to think if there are any other, are there any other tactical questions here that we haven’t covered in terms of how the platform works, how authors can access these deals?

Trip (24:17.248)

Exactly. Yes.

Trip (24:35.212)

I just say we’ve made this very self-serve and very, easy. An author can basically sign up and make the content available in just a matter of a few minutes. We try to make it as fun as using TikTok or something like that, but it’s for authors. So if you’re an author and you’re interested, just go to createdbyhumans.ai, click the Get Started button, it just takes a few minutes to go from there. And if you’re a publisher, please reach out, we’d love to have a conversation.

Ricci (25:00.675)

Okay, great. And then you already have authors, it sounds like, who you have worked with and who have signed some deals. Are there any high profile authors or authors people might know who you can speak to or speak about?

Trip (25:15.182)

Well, we, you can see a lot of our authors are on our website. A couple that we work really closely with who are really involved with the company are Walter Isaacson and Doug Preston. They’ve been kind of, they’re both, you know, investors and they’ve been involved in as thought partners really early on. And yeah, we weren’t close to with them and we’ve got, you know, really nice growing list of authors. can check it out on our website and yeah, we’re going to have some pretty cool things to announce.

pretty soon, which will really kind of demonstrate how their books are being brought to life and being monetized in new ways in an AI world.

Ricci (25:52.173)

Do you think that the rights demand is going to skew more towards nonfiction works or do you think it’s going to be pretty equally distributed between fiction and nonfiction?

Trip (26:04.588)

I think it’s a little skewed toward nonfiction. Having said that, there’s definitely interest in fiction and we’ve got some really interesting fiction deals in the works. So I think that there’s interest across the board.

Ricci (26:18.031)

Let’s end with kind of a zoom out questions. For those authors who are really anti-AI and their core fear is that AI will replace them, that they’re to be competing with AI, what would you say to them to ease those fears?

Trip (26:37.326)

Well, I think if you’re, even if you’re really against AI, we still encourage you to sign up for created by humans. mean, we’re, we’re you’re, you know, one of the options we support is you can just select to opt out of licensing entirely. And if you do that, we will do our best to keep your books away from these AI companies. You know, we’ve, we’ve built a system for, monitoring AI companies for whether they’ve, they’ve used the books or not. And we can, we can, uh, we make that data available to you. And, and, uh, to the extent we can try to keep your books out of the world of AI. Um, so we, so we kind of support.

support authors, whether they’re for or against AI, we want to basically give authors full control as their idea. Having said that, I think that AI is very much a trend that’s here to stay. So I think trying to avoid AI is probably not a great strategy as an author. think what’s better is to learn about it and to engage with it and Create About Humans makes that possible. And I think that if we build out the right framework for this and get the business model right, I actually think this could be

really great for authors and, just be, you know, wonderful moment for authors and a wonderful moment for human creativity and storytelling. And that’s the world we’re, we’re trying to build.

Ricci (27:44.322)

Great, thank you. We’ll leave it there. Thank you so much, Trip, for coming on and talking to us about Created by Humans. It sounds like you’re doing some very interesting work. One last time, if you can just give us the web address for Created by Humans and where people can find you.

Trip (28:02.542)

Yeah, it’s createdbyhumans.ai is the website and you can email me at Trip at createdbyhumans.ai or find me on Twitter or LinkedIn. Yeah, it’s just been great to be here today. Thank you everyone for listening and hope to chat more soon.

Ricci (28:21.729)

Great, thanks so much for coming on trip.

Trip (28:23.906)

Thank you.