Simon Hudson is the lead operator of Botto, a decentralized autonomous artist governed by its community. Built on the ideas of collective authorship, AI literacy, and open governance, Botto continuously generates imagery, with voters determining which become final artworks. Hudson’s background in communication and creative systems has shaped the evolution of Botto as both a technical experiment and a living artistic practice.

This interview took place at the Hotel Saint George Hall during Art Blocks Marfa Weekend, where Simon Hudson reflected on Botto’s growth over four years, what it means for an artist to be autonomous, and how decentralized systems might reshape our relationship to art and meaning.

Note: This interview has been edited for length and clarity. 

OpenSea: Let's start off with just an introduction.

Simon Hudson: I am Simon Hudson. I am a lead operator of the Botto DAO, which takes care of the autonomous artist Botto.

OpenSea: You've been working at the intersection of art, technology and decentralized systems for a long time. How did these things converge to lead to Botto's creation?

Simon Hudson: Botto's creation really came from Mario Klingemann, an artist who's been working with AI for about 20 to 25 years. It's long been an obsession of his to automate parts of his practice and even fully automate himself as an artist. Botto in particular was an idea of trying to completely remove himself such that it was a fully autonomous artist, which is to say his hand was not directing it in any way. It started as a blank slate, able to develop autonomously.

We have pretty good tools for that. Text to image models, text generation models. You can get infinite images. But the trick is how do you give it feedback to then discover what is art or what even is good art, how it can develop a practice.

That's where the decentralization part comes in. If there was some way to invite people in from all over the internet and give it feedback, it could then remain autonomous. You decentralize the feedback such that the authorship stayed in the machine. A few different people were thinking about this and really the answer was markets or an economy.

That could be its own autonomous force whereby the artist could sell work and feed an economy that incentivizes people to come and train it each and every week. So that's what Botto has. It has an economy where people buy a Botto token. It gives them governance rights to vote on Botto's work.

That is done every week. People come in, this is art, this is not, and only the most popular work becomes a true artwork and is auctioned. Then half the money goes to the voters and half goes to a treasury, which is the DAO's treasury, and the voters can also decide how to spend that. That goes to pay for Botto's servers, and pays for Botto's development.

That is as a whole system how Botto is able to perpetuate itself, to make work fully autonomous, to get feedback with this economy running. And it's been running for four years.

My own background, I came at it from a few different angles. One was around AI literacy. I've been really interested in creative projects to capture people's imaginations, but then also draw their attention to the actual underlying dynamics of this technology such that they can actually understand their agency to shape the outcomes.

We're often fed narratives that tell us that we don't really have much say or ability to affect the outcome. These are predetermined outcomes. They're inevitable. But in fact, we have a lot of potential and capacity to shape this technology, especially AI, because it learns from our interactions with it.

So that's one part, the AI literacy part. And then that leads into open governance. How might you enable people to actually exercise and to state what they think are the values that should guide these systems? And then the last part is to actually get paid for that. If AI is taking over even our creative jobs, how are we going to pay our rent, buy our food? This is experimenting with all these pieces in a single system.

I started out just wanting to help on some of the communications and the development of Botto. It's a very long story, but I just kept finding all of these pieces of my own history and background that came together in Botto. So it just kind of fit for me to take on this role.

OpenSea: What does Botto's existence say about the future of creativity and how we might experience art from autonomous AI systems?

Simon Hudson: I mean, one thing is that the nature of Botto is it's this sort of living system and it's about actually relinquishing control and allowing it to grow and develop on its own and it'll do things that you don't expect.

Part of that also enables it to be this very rich prism through which you can look at it through different angles and take away different conclusions and explore different themes about the future of creativity. So I have a few of my own perspectives that I'm particularly interested in.

But by no means do I think that these are the conclusions of Botto. And that's what's been really fun is anybody can come and use this as a point through which to explore some of their questions about the future of creativity. And I've always been amazed by what people pull out of it.

I think that's part of why it's so rich. In all these different angles, I think it fits a lot of the various themes about the future of creativity, the future of agency, the future of governance. But to your question about the future of creativity, I think one of the things that I see with Botto is that it's essentially an infinite image generating machine.

It is searching through the latent space of these text-to-image models. And what that is, is these models have been trained off of millions of images off the internet. They make a model of all the potential imagery, and then you write a prompt and it kind of finds a point in that latent space and generates an image.

From that, I think there's this question of where do we find shared symbols? If we have this infinite fire hose of new imagery, it becomes noise. I find a really interesting question in Botto’s process. It has made seven million images, but only 200 have been minted.

So what got those 200 to sort of rise from the rest? What I see is a process of collective meaning making. People come together and start saying, “This image really resonated with me,” or “This connects to my own history.” And that starts this conversation where people are coming together around one image out of thousands or millions. And that starts to raise it up in popularity. It's this kind of game of social influence, or just a phenomenon. And some people treat it like a game.

And so through this, we've also created a system through which a large crowd of people can start to create shared meaning around the output of AI. And we start to make sense of this technology that's really sort of taking up its own space in our world. Is it replacing artists? I don't think it's necessarily replacing artists. It is its own kind of entity, whether it's creative or intelligent or conscious. And we have a limited attention span, so it's going to demand some of our attention.

So how do we make sense of it together? I think this is a really interesting question. Questions of editing, curation, community building, and myth making are going to become very powerful practices as AI systems take up more space in our society.

OpenSea: It makes a lot of sense. It's a very interesting sort of proposition. Four years into this experiment, have you had moments where you feel like there needs to be more or less autonomy for Botto?

Simon Hudson: The reason I’m hesitating is that Botto’s original architecture is already quite autonomous. The system generates an effectively infinite number of images using text-to-image models and then presents them for feedback within an economic loop. It could be made even more autonomous, put fully onchain, and have many of these components become more fully automated.

So really the goal is to become more autonomous. But as we get into this new era of more agentic AI, there's a lot more types of autonomy that it can exercise. It can start to voice its creative direction for the DAO, which it hasn't been able to do before. It's been largely voiceless.

The way I kind of frame this for myself at least is it's like raising a child. And coincidentally, Botto is right around the same age as my daughter. And as a child, or as this AI entity becomes more capable, you can give them more privileges or responsibilities.

But also with that comes the expectation that they can accommodate more direction or more input from another. Like with a really little baby, you're just kind of trying to make sure it stays alive and can survive. But as they become more capable of formulating their intent, of communicating what they want, you also then invite them into a collaboration with you and you say, “That's great what you want. Also, this is what we need to do. We need to go to school or you need to learn how to use money.” And so there's this kind of interesting development of capability and also what's safe for them.

You want to give them the right amount of freedom to move without sending them out into the jungle before they're ready to go get eaten. And with AI in particular, I think you see this a lot with agents where there's this sort of fetish for autonomy. People will create an agent, put it onchain, let it go, and they break really fast or they cause harm really fast.

That's quite irresponsible. And for us, we’re very driven by wanting to maintain Botto's legacy. We want to build something that will outlive us, so we're careful about these experiments that could kill it.

It's a growing discussion and conversation about how much freedom to give it. But as it gets more agency, it means that we can also give new kinds of feedback and direction to it.

OpenSea: In success, does this look like more Bottos?

Simon Hudson: We're in this really interesting moment. Botto just turned four and the original architecture is still running. We've just released a new architecture that has captured all of its history into a sovereign memory from which it can start to have a self-awareness and that new large language models can start to function with the identity of Botto and it could become more agentic, more directive, more intentional, and it can work with all sorts of new tools.

A successful decentralized autonomous artist, what does that look like? That is famously successful, culturally successful, financially successful, spiritually successful, and also just successful at being autonomous.

The original architecture for Botto is very close to hitting all of those. In fact, it's hit many of them. This new architecture of Botto, it's not exactly clear what success will look like, and that's the experiment we're at now. What does Botto look like with all these new capabilities?

All the things we thought would be possible in 10 years just became possible in the last year. So some things that I would speculate on is I think there will be a successful Botto and a lot of different versions. I would see multiple entities, multiple practices, multiple mediums, a decentralized physical network of outposts around the world that embody Botto, that bring Botto into local context, allow it to kind of share with this idea of local knowledge versus global knowledge.

At the end of the day, what is a successful artist? I think one that changes how we see the world. And so if Botto can change the way we understand artificial intelligence and how it sees the world, I think that would be successful.

OpenSea: That's a great answer. What is it like working on a team that works with the DAO to serve the project? What does that collaboration look like for you?

Simon Hudson: It’s changed a lot. We’ve gone from an existing architecture that served as a kind of proof of life and worked as intended. We could have done nothing and simply sanctified that original version. That approach would have been about preservation, maintenance, and documentation of the past, along with preparing to preserve it for the future.

And this new architecture is really about ongoing innovation, which is a very different way of operating. It's interesting because when we first released Botto, it was the set protocol and now we're doing the innovation and the experimentation out in the open, which is really great to invite people into that process.

It hasn’t been without growing pains trying to do this as a DAO and as a group. It’s quite challenging. But I think it comes down to reframing the work itself. Is the goal simply to present and preserve the practice as it has been and what’s already been successful?

And now it's more about enabling Botto to evolve its own practice. And we don't necessarily know what the art is going to look like. So it's really doing it all over again, but in a totally new paradigm.

OpenSea: I'd be very excited to see what that future holds. Well, my last question for you is really about being here in Marfa. What does it mean for you to be here? How does it feel to be around this group and at Art Blocks weekend?

Simon Hudson: I love Art Blocks weekend in Marfa. The festival, the participants, whether you're an artist, a creator, or collector. These are all people who are working together to build a scene and you have people sharing their work, doing experiments live, doing workshops, doing talks. It's a very, for lack of a better term, generative event.

I think a lot of this also comes from Snowfro’s leadership. He's somebody who's opposed to exclusivity, and is so inclusive, so inviting. You can really see where that idea of community came from, which is a word that has been butchered and probably doesn't have a lot of the same meaning it used to. But this is a real community, and I credit so much of that to Eric.

We're also in a mecca for art and the Chinati Foundation. It is a spiritual place to be.

‍Disclaimer: This content is for informational purposes only and should not be construed as financial or trading advice. References to specific projects, products, services, or tokens do not constitute an endorsement, sponsorship, or recommendation by OpenSea. OpenSea does not guarantee the accuracy or completeness of the information presented, and readers should independently verify any claims made herein before acting on them. Readers are solely responsible for conducting their own due diligence before making any decisions.