How Do We Want AI to Influence Human Creativity?

Alex Reben (0:00)

If you look throughout time, right, there's obviously nothing quite like AI in the past, but I feel like there's a lot of times where technology, even if we see it as kind of primitive back in the day, has changed how we express ourselves, what it kind of means to be human, how it's changed our traditions and what we do. And I think we have decision points at those times to figure out how do we want this new technology to represent us and how do we want to use this to move into this new future.

Laurie Segall (0:26)

I'm Laurie Segall, and you're listening to Mostly Human, a tech podcast through a human lens. Alex, welcome.

Alex Reben (0:34)

Thanks for having me.

Laurie Segall (0:35)

I'm happy to have you here. I think we've been going, there's been a lot of episodes that make us worry about the state of humanity and AI. And so you're coming in at the exact right time where I feel like you're going to give us some optimism.

Alex Reben (0:50)

I hope balanced optimism at least, yeah.

Laurie Segall (0:52)

Some balanced optimism. You're kind of at this intersection of art and entrepreneurship. What's the connection point?

Alex Reben (0:59)

Oh, very interesting. Yeah. So I think a lot of artists are entrepreneurs by nature, right? You're out there trying to get your work in the world, right? So entrepreneurship, I think, and art are connected a lot of times. Specifically, what I was looking to do was take a lot of the ideas I've been working with throughout the years, specifically like moving between the digital to the physical realm, and figure out how that could be expanded to a more useful area.

And the past couple of years, a lot of the technology has come together, where I think we're at a point that things like Agenda AI and the APIs and connections to the physical realm will allow people to go from their imaginations to physical manifestation.

Laurie Segall (1:42)

Yeah.

Alex Reben (1:43)

Which is connected to a lot of what I've been interested in. My background being like in robotics and math as well. So I'd always been interested in like how technology can connect to the physical world, which is all what robotics is.

Laurie Segall (1:53)

Which is super interesting now, right? To think like we are now, we spent the last years really kind of honing in on AI, the conversation around AI. And now one of the biggest things we're hearing is that AI and the physical world is going to become kind of the next frontier, which I definitely want to get into.

But I'm not done with your background yet. So I'd love to talk a little bit more about, so you were the first artist in residence at OpenAI. What is an artist at OpenAI, artist in residence at OpenAI actually do? I mean, and especially at a time where a lot of folks were like, looking at AI is scary and doing a lot of these, an artist walks into OpenAI, what happens?

Alex Reben (2:34)

Yeah. So that kind of came about because I had been working with them for a couple of years. I got early access to GPT before it was ChatGPT and had been making a lot of outputs that people thought were fun and turning that into artwork. And then I had a lot of conversations with them internally. And we were like, hey, wouldn't it be cool if I could kind of come in a bit closer, be closer to folks, be able to talk a bit deeper about things.

As far as what an artist does within a tech company, it varies between different companies. Given OpenAI was kind of, I guess, relatively early at that point.

It was, and I was able to get close to folks.

I was able to make artwork which used technologies in ways that were unexpected and created outputs which were unusual. Like what? Well, so, I mean, the most obvious one was towards the kind of end of that, where I got access to Sora before it was out there, and I saw a bunch of cool videos and how people were playing with it. And I realized that if you instructed Sora to have an object on the screen and rotate it, it actually kept the 3D perspective of that thing. And then there were computational techniques to take 3D models out of video, if you were to walk around a sculpture. That's what I did. I prompted it to basically walk around a imaginary sculpture and I was able to pull out a 3D model of that thing, and then turn it into a real sculpture. And at that point, they weren't really looking at Sora to be like a text to 3D generator. It was more like a text to video type thing. So that was like an unexpected use of the technology to do something different.